link
stringlengths 41
45
| date
stringlengths 9
9
| paper
dict | reviews
listlengths 1
6
| version
int64 1
5
| main
stringlengths 38
42
|
|---|---|---|---|---|---|
https://f1000research.com/articles/5-132/v1
|
03 Feb 16
|
{
"type": "Research Article",
"title": "Normalizing sleep quality disturbed by psychiatric polypharmacy and sleep apnea: a comprehensive patient-centered N-of-1 study",
"authors": [
"Victoria Magnuson",
"Yanpin Wang",
"Nicholas Schork",
"Yanpin Wang"
],
"abstract": "There is a growing interest in personalized and preventive medicine initiatives that leverage serious patient engagement, such as those initiated and pursued among participants in the quantified-self movement. However, many of the self-assessments that result are not rooted in good scientific practices, such as exploiting controls, dose escalation strategies, multiple endpoint monitoring, etc. Areas where individual monitoring and health assessments have great potential involve sleep and behavior, as there are a number of very problematic sleep and behavior-related conditions that are hard to treat without personalization. For example, winter depression or seasonal affective disorder (SAD) is a serious, recurrent, atypical depressive disorder impacting millions each year. In order to prevent yearly recurrence antidepressant drugs are used to prophylactically treat SAD. In turn, these antidepressant drugs can affect sleep patterns, further exacerbating the condition. Because of this, possibly unique combinatorial or ‘polypharmaceutical’ interventions involving sleep aids may be prescribed. However, little research into the effects of such polypharmacy on the long-term sleep quality of treated individuals has been pursued. Employing wireless monitoring in a patient-centered study we sought to gain insight into the influence of polypharmacy on sleep patterns and the optimal course of therapy for an individual being treated for SAD with duloxetine (Cymbalta) and temazepam. We analyzed continuous-time sleep data while dosages and combinations of these agents were varied. We found that the administration of Cymbalta led to an exacerbation of the subject’s symptoms in a statistically significant way. Further, we unmasked and monitored treatment effects on a latent obstructive sleep apnea condition. We argue that such analyses may be necessary to effectively treat individuals with similar overall clinical manifestations and diagnosis, despite their having a unique set of symptoms, genetic profiles and exposure histories. We also consider the limitations of our study and areas for further research.",
"keywords": [
"sleep",
"depression",
"polypharmacy",
"wireless monitoring",
"sleep apnea",
"personalized medicine",
"N-of-1 trials"
],
"content": "Introduction\n\nWinter depression or seasonal affective disorder (SAD) is an atypical depressive disorder that in most cases has onset in fall or winter with remission in spring or summer. It is estimated that approximately 5–10 percent of people in the U.S. (i.e., 10–20 million people) experience varying degrees of SAD in a given year1. While full syndromal SAD (frequently dependent on additional external negative stressors) is not reached every year, subsyndromal symptoms can be seen2. These symptoms are multiple, and include varying degrees of hypersomnia, carbohydrate-craving and jet-lagged physical and mental states (what is known as “brain fog”) resulting in fatigue and irritability. The annual shortening of the photoperiod is believed to be the main factor in SAD onset; however, responses to cold temperatures and epigenetic changes have been documented in seasonal mammals and exhibit evolutionary conservation down to lower forms of life3–6, suggesting that many very basic physiologic mechanisms could contribute to SAD. Ultimately, SAD is a complex disease with both chronobiological and neurobiological underpinnings7–11, which may include an etiology that for some could even begin in utero12–16.\n\nTreating SAD is far from trivial and will require tailoring the treatment to an individual and his or her circumstances, for a whole host of reasons, not the least of which concern both individual and societal expectations regarding work habits, lifestyle, communal conventions surrounding day vs. nighttime activities, and the use of pharmacotherapies to treat conditions affecting behavior. In addition, SAD, and depressive syndromes in general, are known to be accompanied by many co-morbidities and sequelae, including anxiety, detrimental body habitus, anhedonia, and, more importantly, sleep disturbances which may exacerbate any underlying depression as well as the additional associated conditions2. Tailored treatments for each and every condition possessed by an individual patient who also has SAD could adversely affect that patient’s sleep, thereby creating negative feedback for the SAD-related and other symptoms. Treatment of SAD includes a general recommendation for morning bright light therapy and/or antidepressant treatment which can be somewhat effective in managing symptoms, while melatonin, exercise and negative ion therapy are also suggested. However, a recent critical review of light therapy literature showed that most bright light therapy studies have methodological issues and evidence is not unequivocal17. Further, cognitive response to bright light therapy can vary based on genetics18. A proper prescription for light therapy requires knowing the dim light melatonin onset (DLMO) of SAD individuals (2/3 are phase-delayed) to determine circadian phase19. The same is true for using supplemental melatonin to advance sleep phase, as improper timing and dosing can exacerbate symptoms19. Because of the seasonal “on-off” nature of the disorder and difficulty in long-term compliance with bright light therapy (due to eyestrain and lack of individualized prescription), year-round prophylactic treatment with antidepressants may be prescribed.\n\nTreatment for SAD and its sequelae are also compounded for peri- and post-menopausal females – a fact which may be under-appreciated in the primary care setting. The progression to menopause in normal women can result in circadian rhythm, vasomotor, and sleep disturbances and an increased risk for depression, possibly further exacerbating symptoms20–22. Therefore, a clinician’s choice to potentially increase the dosage of, e.g., a previously effective SSRI antidepressant can in turn exacerbate side effects, such as sleep disturbances. Importantly, sleep apnea is one of the most under-diagnosed conditions in post-menopausal women and is a leading cause of cardiovascular morbidity and mortality23–29. Prescribing sleep medications to aid in depression-related symptoms in peri- or post-menopausal women that may be susceptible, or have, sleep apnea is therefore highly problematic.\n\nThe fact that depression and sleep disturbances go hand in hand thus creates even more difficult treatment challenges. For example, ironically, it is known that many first-generation antidepressants exert their effects by, among other things, restoring sleep. Unfortunately, many second-generation antidepressants disrupt sleep. It is now accepted that SSRIs and SNRIs typically used to treat SAD can cause sleep disturbances, both in sleep quality (sleep initiation and maintenance) and sleep architecture (rapid eye movement (REM) and non-REM (NREM) sleep)30–35. Further, these agents can induce or escalate parasomnias such as periodic leg movements (PLMs) and restless legs syndrome (RLS)36,37. These effects on sleep could further lead clinicians to routinely prescribe sleep medications to counter the stimulating effects of antidepressants, as was recommended for insomnia in patients taking fluoxetine38–40. However, sleep medications can have their own negative impacts on sleep quality and architecture, and are not recommended for maintenance use. Thus, the resulting polypharmacy used to treat SAD is usually pursued without regard to the timing or dosage of the drugs or concern for drug-drug interactions. This fact, combined with unique patient characteristics such as age, gender, genetic and exposure profile, and co-morbid conditions, can further impact response to any prescribed drug or drug combination and may change over time.\n\nIn order to combat these issues, the management of SAD and related psychiatric disorders should, as noted, be pursued in a more patient-specific or ‘personalized’ manner – something that might not be accomplished at the level of a primary care provider. How such personalization can be achieved generally is an open question given the costs associated with the extra time a clinician might have to spend with a patient to determine an optimal course of therapy, but does suggest a greater number of empirical studies investigating the effects of polypharmacy and the utility of different treatment strategies are needed. In addition, patient-acceptance of the challenges surrounding treatment may motivate self-assessments of the type being pursued by members of the quantified self movement but perhaps in more objective ‘N-of-1’ clinical trial like settings41,42. We describe a study investigating the influence of polypharmacy involving a 58-year-old post-menopausal female who was diagnosed with SAD in 2001. The ultimate goals of the study were two-fold: to determine if objective claims about the influence of her treatments on her psychological well-being could be made in a self-assessment-oriented but designed N-of-1 study, and whether her medication use correlated with exacerbation of her various symptoms and conditions. It should be noted that consistent with difficulties in treating SAD discussed above, the patient had a history of clinic visits, switching medications, multiple symptoms and general dissatisfaction with the medications she was taking.\n\nUltimately, the study leveraged wireless monitoring devices and regression modeling to assess patient sleep quality (e.g., the Zeo Sleep Monitor43,44), and designed a drug removal and dose escalation study to determine drug effects. In the course of the study, a number of important insights were obtained. For example, an abnormal sleep architecture was observed, seemingly caused by her use of temazepam. Once the sleep aid temazepam was removed, an increase in deep sleep was observed. However, a Cymbalta-induced disruption of sleep initiation and REM latency remained. Decreasing the dose from 60 mg to 30 mg for Cymbalta did not appreciably change the outcome. Because the subject was experiencing continued sleep difficulties and given her co-morbid seasonal circadian rhythm disorder, more sophisticated wireless monitoring was initiated, including actigraphy, heart rate monitoring, breath rate monitoring, temperature collections, and PLM detection.\n\nInterestingly, removal of Cymbalta completely restored sleep initiation and decreased REM latency. In addition, Cymbalta caused an increased rate of PLMs compared to periods in which Cymbalta was not used. When the subject was taken off all medication completely, normal sleep architecture was restored, but sleep apnea was observed. In light of this, monitoring of the subject’s change in heart rate variability (HRV) during a clinical intervention for obstructive sleep apnea using advancement of a mandibular splint was pursued and a consistent pattern indicative of an underlying pathology was observed. The study identified a number of statistically significant correlations between medication use and symptomology that led to a number of potential recommendations for future treatments. Although it is important to acknowledge the shortcomings of the study, we feel that such patient-engaged and initiated yet protocol-oriented and designed N-of-1 studies may be the best way to individualize treatments for individuals with multiple mood and sleep-related conditions for which polypharmaceutical interventions are common.\n\n\nMethods\n\nWe studied a post-menopausal 58-year-old female (the ‘subject’, author VLM) interested in self-monitoring and an N-of-1 study for her sleep disturbances given her lengthy dissatisfaction with available treatment options, lack of insights into her multiple conditions, and a very elaborate and complex treatment history. The subject had a long history of usage of benzodiazepine as a sleep medication while taking antidepressants. The subject had a sleep study performed in 2003 to rule out sleep apnea. However, during this sleep study, the subject was given zolpidem (Ambien), but reportedly did not sleep well in the noisy hospital room setting possibly confounding the detection of sleep apnea. In fact, the sleep study report showed a high rate of PLMs, little if no deep sleep, delayed REM sleep, and was ruled negative for sleep apnea. The subject self-reported evening tendencies, trouble sleeping and easily jet-lagged. She noted that sleep deprivation during the week resulted in a sleep pressure made up for on weekends. The subject reported that she felt she performed best after 10 am and 8–9 hours sleep, and had a preferred sleep/wake schedule of 11–11:30pm/7–8am. The subject has relatively low blood pressure, but has taken supplemental magnesium before bed, which she feels helps avoid restless legs. The subject loosely qualifies as evening prone or delayed sleep phase disorder according to Basic Language Morningness Scale (BALM) questionnaire, which uses a 6-item scale45. In 2001, the subject had been diagnosed with seasonal affective disorder (SAD) during what she perceived as a stressful, cloudy winter while living in upstate NY. As a result she was prescribed 40 mg fluoxetine (Prozac) daily year-round while residing in winter climates, and she also has used light boxes during winter months at home and at work. In July 2006, the subject moved to San Diego, CA and at that time was taking 40 mg fluoxetine (Prozac) daily plus 50 mg sumatriptan (Imitrex) for morning headaches, as needed. In August 2006, the subject was switched to generic 40 mg fluoxetine and prescribed 0.5 mg alprazolam for sleep. In March 2007, she was switched to 1 mg clonazepam, indicated for her PLMs. In winter 2008, at age 54, the subject switched from 40 mg fluoxetine to 60 mg Cymbalta. Despite the use of sleep medication, the subject continued to have sleep-disordered breathing and PLMs, as documented by her husband. In 2011, the subject requested 10 mg zolpidem as an alternative sleep medication, since after becoming post-menopausal the subject felt that her sleep quality was decreasing. Notably, while in San Diego, her sumatriptan usage increased in frequency and to a 100 mg dose which was needed to alleviate morning headaches. Finally, in summer 2012, she reported that under prolonged indoor low-light conditions she was susceptible to feeling fatigued, exhibiting seasonal symptomology even in summer months in San Diego. In fall 2012, the subject was taking 60 mg Cymbalta, 30 mg temazepam for sleep, and 100 mg sumatriptan as needed for morning headaches. Note that during the course of this patient history narrative, originally prescribed brand medications were eventually replaced with generics. Given this elaborate history of polypharmacy, multiple diagnoses, general dissatisfaction with her treatments, and desire to be studied to potentially enable a better treatment strategy, an N-of-1 study was pursued to explore how her medications affected her sleep in the context of her diagnosed winter depression (SAD), evening chronotype, delayed sleep phase, restless legs/PLMs and morning headaches.\n\nThe present study was self-administered by one of the authors (VLM). Therefore, ethical approval from an Institutional Review Board was not sought because the Helsinki Declaration does not apply in this case.\n\nSleep and activity monitoring. To assess sleep patterns a Zeo Sleep Monitor (http://www.myzeo.com, model number ZEO 301) was used, which was worn nightly after entering bed per manufacturer instructions. The Zeo wirelessly tracks sleep stages at 5-minute intervals and has been validated against laboratory polysomnography43. The number of awakenings (after sleep onset), percent time in light, deep, REM and wake were recorded and assessed with an accompanying iPAD application (Zeo Sleep Manager v1.9.0). Until the manufacturer’s bankruptcy, the Zeo online application provided nightly tracking of sleep stages and tools for evaluating trends. In addition, educational materials reminding the user of good sleep hygiene practices and journaling and counseling options were also offered. The data obtained with the Zeo monitor was captured on an iPad and Zeo graphic image data obtained with the device is available from the authors. In addition to the Zeo monitor, an Actiwatch Spectrum (manufactured by Philips Respironics) was used to collect data at 15-second intervals and worn daily to track sleep and light exposure. It was synchronized to the Zeo monitor on the nights it was worn. Because Actiwatch relies on movement to score wake versus sleep, the Actiwatch tends to overestimate time in sleep and underestimate time resting in a quiet awake state (Actiware software version 04.00). Periodic leg movements were measured using the PAM-RL (also manufactured by Philips Respironics) right and left ankle sensors and scored using default settings in software (PAM-RL version 7.6.2). Finally, the Fitbit Ultra actigraphic monitor (http://www.fitbit.com) was worn daily to track walking or “step” activity. The subject wore the Fitbit on her waist from the start of her day through the evening. The Fitbit can be used to monitor sleep activity, but may overestimate sleep time since it keys off of movement (Fitbit app v1.8.2).\n\nVital signs. The Equivital belt (Hidalgo, belt type EQ02) was used to collect data at 15-second intervals and worn nightly to measure heart rate, breathing rate, and skin temperature. The device has been shown to be reliable for heart rate and R-R interval measures during sleep based on Hidalgo data analysis software quality measures (Equivital software, EQ Manager version 1.1.29.3883). However, clear movement artifacts (R-R interval spikes) could be identified and were removed from the data (see the Procedures section below). Unfortunately the belt was less reliable for breathing rates possibly due to stretch sensor impingement when lying down. While nightly traces showed periods of normal breathing, they also showed long periods (20 minutes) of zero breathing, scored as apnea by software indicating breaths below the limits of detection. Similarly, skin temperature could not be reliably tracked for circadian purposes, due to the side pocket sensor design that compromised temperature readings while lying on the left side of belt. The belt may not be ideal for the female anatomy given how it is to be worn. As a result, although not all of the breathing rate data could be reliably used across the entire study, we were able to uncover data patterns (especially when timepoints were overlayed across the other devices) that were indicative of sleep apnea/hypopnea and led us to the subject being clinically tested.\n\nPharmacotherapy manipulation: effect on sleep. A schedule was developed for evaluating the effects of Cymbalta, temazepam and melatonin on the subject. Essentially, Cymbalta, temazepam and melatonin (Nature Made, 3 mg chocolate melts) were provided to the subject in pre-specified time periods with pre-specified doses initiated on weekends. Fourteen trials were conducted from 12-30-2012 through 07-05-2013. Description of the 14 trials and the number of nights with complete data are listed in the Table 1 (abbreviations: Cymbalta (CYM); temazepam (TEM); melatonin (MEL)). Melatonin was used to attempt to phase-shift the subject as needed to keep a work schedule, but several periods involving different combinations were pursued to explore the influence of melatonin on phase. It should be noted that in designing a study like the one described there are a number of potential confounding variables that inevitably arise in any naturalistic, free-living setting assessing sleep quality: a) sleep consolidation could occur as sleep deprivation leads to sleep pressure as week progresses; b) sleeping in and changing sleep patterns on weekends could affect weekday trends; and c) percent time in wake after sleep onset can be increased by PLMs, sleep apnea or other sleep maintenance problems, which could be compounded by medication use.\n\nSleep analysis. Each night and morning, the subject manually entered start and stop times into the Zeo sleep monitor iPAD app. The time to REM sleep was manually calculated based on Zeo graphic histogram output showing first REM sleep bar. Percent wake, light, deep and REM sleep and number of awakenings were supplied by the Zeo device. We did not use the Zeo sleep latency parameter “Time to Z” due to the confounding presence of PLMs, which our subject has shown to exhibit upon sleep initiation (clinically validated via videotape). The subject also wore the Actiwatch Spectrum around the clock from April 2013 until August 2013 as well as the Equivital belt and PAM-RL ankle sensors nightly from April 2013 to July 2013. Some missing sleep quality data occurred due to days for which the subject was traveling or when she was required to wear alternate devices for sleep apnea diagnoses.\n\nSleep apnea treatment: effect on HRV. After clinical diagnosis of obstructive sleep apnea in August 2013, the subject was fitted for a mandibular splint or mouthguard (MG). After fabrication and refitting adjustments, the device was ready to be worn nightly in late October 2013. Following her physician’s schedule of treatment, the subject wore the MG to treat sleep apnea. The subject was also periodically monitored by her sleep physician with a Sleep Image device (http://www.sleepimage.com), a 2-electrode device which measures cardio-pulmonary coupling during NREM to determine progress during treatment. The unadjusted MG was denoted as MG0x, but simply inserting a mouthguard creates vertical displacement and also (by design) some horizontal displacement of the jaw. Ultimately 3 MG adjustments (zero, four or six turns of the splint armature; denoted as 0x, 4x, and 6x) were made to the device. As noted below, we also monitored the subject with Zeo, Actiwatch Spectrum and Equivital devices from November 10, 2013 to December 19, 2013. The number of Zeo measured 5-minute sleep bouts for every sleep stage in each night was recorded. Both sleep apnea and treatment for sleep apnea can obviously compromise sleep quality and were important to accommodate and assess.\n\nHeart rate variability analysis. Heart rate was assessed with a number of devices. Equivital R-R interval data was collected and after movement artifacts were removed R-R interval nightly averages and R-R interval standard deviations were calculated. Movement artifacts were defined as R-R data spikes < 500 ms and >1100 ms and the artifact data was imputed by filling in the preceding value. Respiratory sinus arrhythmia is a coupling of the heart rate and breath rate (cardio-pulmonary coupling, CPC). Based on coupled autonomic-respiratory oscillations, “stable” sleep shows high frequency coupling (HFC), “unstable” sleep shows low frequency coupling (LFC), while wake and REM sleep show very low frequency coupling (VLFC)46. Therefore, measures of CPC to detect elevated LFC are usually collected during NREM (light + deep) sleep, which occurs mainly during the first half of the night. In order to accommodate this biological phenomenon and in an attempt to equalize the amount of the data generated by the 15-second Equivital heart rate collection rate, we used an “end-of-night” truncation of 6 am. The collection period used for data analysis was from first sleep bout to last sleep bout before 6 am. HRV was defined using the standard deviation of the R-R interval (SDRR), also referred to as the SDNN method (standard deviation of normal-to-normal beats) used by others47,48. Actiwatch-defined sleep intervals and Zeo-defined first sleep bouts were used to define beginning of sleep and end of night. This was done to normalize behavior after sleep onset. Given that HRV varies with sleep stage, sleep stage transition and time of night, it was important to define an interval that began with sleep onset and ended at the same time every morning.\n\nAll analyses were performed using R version 3.1.3 (http://www.R-project.org). For the sleep analysis, the data used contained information for 188 consecutive nights from December 30, 2012 to July 5, 2013 with 21 nights having missing data attributable to lost records and was therefore treated as missing at random (MAR). The response variables focusing on sleep quality included the number of wakes, time to first REM sleep, percent time in REM sleep, percent time in deep sleep, percent time in light sleep, and percent time in wake. To accommodate the presence of serial correlation in the nightly data, linear models considering an autoregressive moving average (ARMA) serial correlation structure among the data were fit. Different assumptions about the degree of serial correlation were made and tested. Interestingly, little evidence for a strong serial correlation was found, and therefore simple univariate linear regressions were used for all response variables via the lm function in R, retaining predictor variables significant at p < 0.05. Analyses involving model residuals were pursued to assess goodness-of-fit and satisfaction of linear model criteria. These included a Durbin-Watson test (to detect serial correlation between residual values), Shapiro-Wilk normality check, Portmanteau test and ARCH test. In cases where residuals in final models did not satisfy normality, a Box-Cox procedure was performed on the model. The resulting optimal exponential transformation was applied to the response variable and the model refit. To determine best fit among similar models, linear regression model fit measures (Akaike information criteria (AIC), Bayesian information criteria (BIC) and log likelihood) were evaluated. Only the best final models meeting all linear model criteria including no serial correlation or autocorrelation are presented in the results. The univariate regression models for each dependent variable were pursued in very similar ways, as outlined in the following example. Let perstaget denote series analysis response variables, where non-transformed variables are percent wake (perwake), percent light (perlight), percent deep (perdeep) and percent REM (perrem).\n\nTo be more specific, an example model for perstaget was created to follow the simple scheme below, with other variables leveraging similar models:\n\nperstaget = μ0 + βcym30 ∗ cym30 + βcym60 ∗ cym60 + βmel3 ∗ mel3 + βcym30mel3 ∗ cym30mel3 + βcym30mel6 ∗ cym30mel6 + βcym60mel3 ∗ cym60mel3 + βcym60mel6 ∗ cym60mel6 + βcym60tem15 ∗ cym60tem15 + βcym60tem30 ∗ cym60tem30 + ∈t\n\nwhere μ0 is a y-intercept term, the β terms are regression coefficients, ∈t is an error term with 0 mean and variance σ2. The other terms in the model correspond to the drugs being evaluated and are denoted as follows: Cymbalta 30 mg (cym30); Cymbalta 60 mg (cym60); Melatonin 3 mg (mel3); Cymbalta 30 mg and Melatonin 3 mg (cym30mel3); Cymbalta 30 mg and Melatonin 6 mg (cym30mel6); Cymbalta 60 mg and Melatonin 3 mg (cym60mel3); Cymbalta 60 mg and Melatonin 6 mg (cym60mel6); Cymbalta 60 mg and Temazepam 15 mg (cym60tem15); Cymbalta 60 mg and Temazepam 30 mg (cym60tem30). Significant terms (i.e., p < 0.05 based on t-test of the coefficient value and its standard error) in the model were evaluated in an overall model fit as well as in a step-wise manner. Models were also fit to assess the impact of study design (night in time course) and days of the week (using Sunday as comparator per convention) by including these factors as independent variables in the model. The same analyses were performed for time to REM sleep.\n\nFor heart rate variability analysis, the data were collected for 40 consecutive nights beginning from November 10, 2013 to December 19, 2013. Four days of missing data were due to Zeo equipment malfunction or a need to wear alternate head devices. We treated missing data in these analyses as missing as random (MAR). As noted, during this time, the subject wore the mouth guard (MG) for 33 days with settings 0x (13 days), 4x (11 days), and 6x (9 days). The R-R interval data was approximately normal within each MG setting.\n\nVariable selection for R-R interval standard deviation (measure of HRV) was performed starting from a full model which included MG setting (0x, 4x, 6x) and manually dropping terms with p-value greater than 0.05. Models were also tested for the impact of study design, days of the week, and model residual diagnosis was assessed as described above.\n\n\nResults\n\nSleep data was collected for 188 consecutive nights from December 30, 2012 to July 5, 2013, with 21 nights having missing data. Table 2 gives a descriptive analysis of the sleep parameters used in our study. The mean and standard deviation (SD) for: the number of times per night the subject was awakened (wakes (N)); time to first REM sleep bout in hours (1st REM (h)); and percentage of time in each sleep stage (wake (%), light (%), deep (%), REM (%)) at each drug dose is shown. The number of days per dose and percent of the total nights are also shown (N days (%)). The dataset was not balanced in the sense that we had different numbers of observations while the subject was on different dosages of a drug.\n\nFigure 1, Figure 2 and Figure S1–Figure S4 (see Supplementary Material) graphically depict the impact of Cymbalta, melatonin and temazepam drug use on the subject’s sleep architecture. Figure 1 and Figure 2 show the percent of time per night that the subject was in deep sleep and light sleep, respectively, during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Similar figures for the number of times the subject was awakened, time to REM sleep, percent time after sleep onset that the subject was awake and percent time in REM sleep during 5-minute intervals detected by the Zeo Sleep Monitor are presented in the Supplementary Material (Figure S1–Figure S4, respectively).\n\nThe percent time subject was in deep sleep during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A–N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nThe percent time subject was in light sleep during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A-N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nA clear relationship can be seen between temazepam intake and reduced deep sleep in favor of light sleep (Figure 1 and Figure 2). However, Cymbalta had the strongest impact on the subject’s sleep architecture as shown in Figure 3, Figure 4, and Figure 5. Cymbalta intake increased the number of awakenings (Figure 3), time to first REM sleep (Figure 4), percent time after sleep onset that subject was awake (wake) (Figure 5A) and in light sleep (Figure 5B) at the expense of deep (Figure 5C) and REM (Figure 5D) sleep. Removal of Cymbalta decreased the number of awakenings, time to first REM sleep, percent time in wake and light sleep and increased percent time in deep and REM sleep (Figure 3–Figure 5).\n\nThe number of times per night subject was awake during 5-minute intervals detected by the Zeo Sleep Monitor. Doses of Cymbalta were decreased from 60 mg to 0 mg.\n\nThe number of hours (h) per night before subject achieved first REM sleep bout during 5-minute intervals detected by the Zeo Sleep Monitor. Doses of Cymbalta were decreased from 60 mg to 0 mg.\n\nPercent time after sleep onset subject was awake (A); subject was in light sleep (B); subject was in deep sleep (C); or subject was in REM sleep (D) during 5-minute intervals detected by the Zeo Sleep Monitor. Doses of Cymbalta were decreased from 60 mg to 0 mg.\n\nBecause of the free-living nature of our study, the subject’s polypharmacy and struggle to counter sleep disturbances, a large variability in the data is seen. In addition, “normal” sleep staging typically follows a pattern wherein the first non-REM sleep (light plus deep sleep) and REM sleep cycle is completed in 70 to 100 minutes, followed by 90 to 120 minute cycles, with deep sleep bouts gradually disappearing and REM sleep bouts lengthening throughout the night49. Near the end of the night, usually only light and REM sleep periods make up the sleep cycles. As a result, we chose to analyze the percentage of time the subject was in each sleep/wake state, rather than total time. For the purposes of comparing Zeo monitored stages to classically defined sleep stages, we assumed the following to represent approximately “normal” sleep stage percentages: wake 5 percent; light 45–55 percent; deep 20–25 percent; REM 25 percent49.\n\nTable 3 summarizes the results of our univariate analyses when the sleep stages, wake, light, deep and REM, were taken as dependent variables. The univariate linear regression models were performed as described (see Methods) and data is presented as mean percent for each sleep stage with treatment effects adjusted relative to the intercept. Analyses of percent wake and light sleep met Durbin-Watson test criteria once two outlier nights each were removed. Final model diagnosis showed that all linear regression assumption requirements were satisfied except for the normality condition for percent wake and percent light sleep. Therefore, the Box-Cox procedure and transformations were performed and the models refit. Final models satisfied all diagnostic tests and the transformed mean estimate values (denoted as ‘bc’) presented in Table 3 were adjusted and back-transformed to give mean percent wake and light sleep.\n\nAdjusted mean in percent, mean estimate or transformed mean estimatebc, standard error (SE) and p-value (Pr > |t-value|). R2: R-squared; bc: Box-Cox transformed variable raised to exponent given in final model. Back-transformation to original units was performed (after adjustments relative to intercept) by taking the nth (exponent) root of estimate.\n\nFrom Table 3 it is clear many of the drugs, doses and drug combinations have a highly significant and negative impact on deep and REM sleep (with the exception of melatonin at 3 mg). The estimate of the y-intercept (μ0) for the model with deep sleep as the dependent variable suggests that approximately 22.3 percent of the time the subject was in deep sleep without any drug effects (p < 2×10-16). The estimated coefficients for the drug and drug dosage independent variables in the model provide the effect on deep sleep of the drugs. The mean percent deep sleep ranged from 4.5 percent (-0.18 (SE: 0.01), p < 2×10-16) while the subject was taking 60 mg Cymbalta and 30 mg temazepam to 18.7 percent (-0.04 (SE: 0.01), p = 0.0068) while the subject was taking 3 mg melatonin. Although temazepam dosing in combination with Cymbalta had the greatest negative impact on deep sleep in favor of light sleep, Cymbalta alone continued to interfere with deep sleep.\n\nSimilarly, the impact of an antidepressant such as Cymbalta is expected to show a decrease in REM sleep, mainly through the delay in REM sleep onset (see Table 4). The estimate of the y-intercept (μ0) for the model with REM sleep as the dependent variable suggests that approximately 34.2 percent of the time the subject was in REM sleep without any drug effects (p < 2×10-16), which might be considered high compared to the usual 25 percent. The estimated coefficients for the drug and drug dosage independent variables in the model provide the effect on REM sleep of the drugs. The mean percent REM sleep ranged from 15.8 percent (-0.18 (SE: 0.02), p = 1×10-14) while the subject was taking 60 mg Cymbalta and 15 mg temazepam to 24.8 percent (-0.09 (SE: 0.01), p = 5.7×10-11) while the subject was taking 30 mg Cymbalta. Interestingly, there was an increase in REM sleep on Thursday, Friday and especially significant on Saturday (39.6 percent (0.05 (SE: 0.01), p = 3.1×10-5)).\n\nAdjusted mean in hours (h), transformed mean estimatebc, standard error (SE) and p-value (Pr > |t-value|). R2: R-squared; bc: Box-Cox transformed variable raised to exponent given in final model. Back-transformation to original units was performed (after adjustments relative to intercept) by taking the nth (exponent) root of estimate.\n\nMost drug combinations, except melatonin, significantly increased time in wake and light sleep. Of note, the Zeo monitor can detect micro-arousals as well as conscious wakes. Thus, some scores of the wakes at night may actually be classified as light sleep. However, from Table 2 and Table 3 the drug combinations increase both of these at the expense of deep and REM. The effect of increasing light sleep at the expense of deep sleep is most notably seen with temazepam use. The estimate of the y-intercept (μ0) for the model with light sleep as the dependent variable suggests that approximately 35.4 percent of the time the subject was in light sleep without any drug effects (p < 2×10-16). The estimated coefficients for the drug and drug dosage independent variables in the model provide the effect on light sleep of the drugs. The mean percent light sleep ranged from 38.4 percent (0.02 (SE: 0.01), p = 0.0201) while the subject was taking 30 mg Cymbalta and 3 mg melatonin to 58.9 percent (0.23 (SE: 0.02), p < 2×10-16) while the subject was taking 60 mg Cymbalta and 30 mg temazepam.\n\nThe major impact on wake after sleep onset occurred after the removal of temazepam and during Cymbalta use, indicating a possible sleep maintenance issue. The estimate of the y-intercept (μ0) for the model with wake as the dependent variable suggests that approximately 10.5 percent of the time the subject was in wake without any drug effects (p < 2×10-16). The estimated coefficients for the drug and drug dosage independent variables in the model provide the effect on wake of the drugs. The mean percent wake ranged from 17.8 percent (0.10 (SE: 0.03), p = 0.0034) while the subject was taking 60 mg Cymbalta and 30 mg temazepam to 31.0 percent (0.22 (SE: 0.04), p = 1.3×10-6) while the subject was taking 60 mg Cymbalta and 3 mg melatonin. Interestingly, there is evidence for decreased time classified as wake as the week progresses that might be attributed to a number of things such as increasing sleep pressure during the week, relaxed frame of mind and sleeping in on the weekend. In fact, the decrease in wake to 4.8 percent on Saturday seems to approximately parallel the increase in REM sleep on Saturday (approximately 5 percent) with similar p-values. There was no impact of the night of the study on any of the models.\n\nTable 4 shows the univariate analyses of time to REM sleep in hours as a dependent variable. The univariate linear regression model exhibited no serial correlation based on the Durbin-Watson test once two Zeo technical outlier nights were removed (known REML error43). As above, models were also tested for the impact of study design (night in time course) and day of the week. An assessment of the normality and serial correlation among the residuals obtained from the model was performed by Portmanteau test, Durbin-Watson statistic, a standard normality check and ARCH test which showed that all linear regression assumption requirements were satisfied except normality. Therefore, the Box-Cox procedure and transformation was performed, and model refit as above. The mean estimates presented in Table 4 were adjusted and back-transformed to give the original unit of hours.\n\nAll drug combinations except for melatonin at the 3 mg dose caused large and highly significant increases in time to first REM sleep. Under normal circumstances the first REM bout is expected to occur before completing the first 70–100 minute full cycle of sleep (light + deep + REM), that is, in less than 2 hours. The estimate of the y-intercept (μ0) for the model with time to first REM sleep as the dependent variable suggests that time to first REM sleep for the subject was 1.27 hours (76.2 minutes) without any drug effects (p < 2×10-16), which is in the correct range for the first full sleep cycle. The estimated coefficients for the drug and drug dosage independent variables in the model provide the effect on time to REM sleep of the drugs. The drug effects ranging from most to least deleterious impact on mean percent time to REM sleep are:Cymbalta 60 mg and melatonin 6 mg, 5.16h (0.76 (SE: 0.08), (p = 4.9×10-16); Cymbalta 60 mg and temazepam 30 mg, 4.43h (0.64 (SE: 0.06), (p < 2×10-16); Cymbalta 60 mg and melatonin 3 mg, 4.33h (0.64 (SE: 0.08), (p = 1.8×10-14); Cymbalta 60 mg and temazepam 15 mg, 4.27h (0.63 (SE: 0.07), (p < 2×10-16); Cymbalta 60 mg, 4.22h (0.62 (SE: 0.04), (p < 2×10-16); Cymbalta 30 mg, 3.73h (0.54 (SE: 0.04), (p < 2×10-16); Cymbalta 30 mg and melatonin 3 mg, 3.55h (0.51 (SE: 0.04), (p < 2×10-16) and Cymbalta 30 mg and melatonin 6 mg, 3.43h (0.49 (SE: 0.06), (p = 8.1×10-15). There was no impact of the night of the study on the model. Of note, is the decrease in time to REM sleep on weekend nights.\n\nThe data shows an unequivocal Cymbalta dose-response, decreasing the time to REM sleep with decreasing Cymbalta dose as expected. Even under the least damaging drug regimen, time to first REM sleep was still delayed over 1.75 hours compared to the maximum in normal sleep architecture (3.43 hours versus 1.67 hours or 100 minutes). This delay in first REM sleep could possibly push normal REM sleep cycling into later parts of the night and interfere with the ability to naturally wake the next morning. Further, truncating REM sleep while keeping a daily work-week schedule might be expected to have additional functional and metabolic consequences.\n\nWe used the data to attempt to predict a lower Cymbalta drug dose which might not be expected to interfere with our subject’s sleep or perhaps normalize all of the percent sleep stages toward “normal” ranges (i.e., wake 5 percent; light 45–55 percent; deep 20–25 percent; REM 25 percent49) since our subject has increased REM (34 percent) and decreased light (35 percent) sleep if drug effects were accounted for. Table 5 provides the predicted values for 10 mg and 20 mg Cymbalta doses based on the fitted regression models.\n\nWe note that even considering the removal of Cymbalta altogether, the percentage of the sleep time our subject was estimated to be in a ‘wake’ period as detected by the Zeo monitor is high. PLMs that tracked with Cymbalta use did decrease to less than 15 per hour during the study (see Dataset 3) which is considered to be normal and therefore not likely to be a source of confusion for the Zeo monitor since episodes of PLMs may confound time in the wake period. However, micro-arousals and unconscious wakes due to sleep apnea in our subject remained a concern and could be reflected in the sleep values we observed.\n\nClinical evaluations suggested that the subject has a deviated septum, a small jaw with substantial retrognathia (overbite), and evidence of clenching and grinding her teeth during sleep. The subject was diagnosed with mild obstructive sleep apnea in August 2013 and a mandibular splint (MG) was fabricated as an intervention. The subject then wore the device as ordered by her physician. After an initial adjustment period, the subject was monitored while the splint was advanced to achieve relief of apnea symptoms (mainly snoring). The monitoring took place for 40 nights from 11-10-2013 through 12-19-2013. The unadjusted MG was denoted as MG0x for analysis purposes. Simply inserting a mouth guard creates vertical displacement of the mouth and jaw and also (by design) some horizontal displacement of the jaw. Subsequent adjustments of 4 turns and 6 turns of screws to advance the jaw were denoted as MG4x and MG6x, respectively. Four nights during the time the MG was considered in the analyses had missing data, three nights had no MG wear and 33 nights had the MG worn at settings 0x (13 days), 4x (11 days), and 6x (9 days) and one outlier night was removed.\n\nWe analyzed heart rate variability (HRV) (RR-STDEV, see Methods) across the entire night and because HRV might be expected to differ among the different sleep stages, we also analyzed the impact of the MG by sleep stage (see Supplementary Material Table S1). We also determined the impact that treating sleep apnea had on our subject’s sleep quality (see Supplementary Material Table S2). A detailed description of these analyses, assumptions and data adjustments made to allow for sleep results comparisons between the pharmaceutical intervention ending in July 2013 and the sleep apnea intervention ending in December 2013 are found in the Supplementary Material.\n\nBy the end of the sleep apnea intervention, the subject’s mean nightly HRV was increased 13.87 milliseconds and similar increases were found during wake, light and deep sleep stages (Supplementary Table S1). The dramatic decrease in wake (apnea wake or micro-arousals perceived by Zeo monitor as wake) and increased light sleep during MG usage suggests an expected physiological response (Supplementary Table S2). Mean percent wake steadily decreased from approximately 15.6 percent (0.52 (SE: 0.04), p < 1.9×10-14) while the subject wore no mouth guard to 2.7 percent (-0.24 (SE: 0.06), p = 0.0004) while the subject wore the mouth guard at setting MG6x. Night in the study did have an effect on wake and was included in the final model. Mean percent light sleep increased from approximately 33.3 percent (0.19 (SE: 0.01), p < 2×10-16) while the subject wore no mouth guard to 39 percent (0.05 (SE: 0.02), p = 0.0185) while the subject wore the mouth guard at setting MG6x. Interestingly, we were unable to detect any significant effect of MG use on percent deep or percent REM sleep.\n\nA clinical sleep study (polysomnography) in January 2014 confirmed the subject to be apnea-free. Ultimately, the final sleep ratios for our subject were “normalized” (wake 2.7 percent; light 39 percent; deep 23 percent; REM 29 percent) by the end of the study (typical values are: wake 5 percent; light 45–55 percent; deep 20–25 percent; REM 25 percent). Note that the final combined average sleep ratios do not quite add up to 100 percent presumably due to errors in estimates for wake and light sleep combined with non-model based values for deep and REM sleep.\n\n\nDiscussion\n\nWe have shown that monitoring an individual’s response to various drugs used to treat her severe sleep and sleep-related disturbances yielded important and actionable insights. For example, the subject’s sleep quality was highly compromised when taking Cymbalta at therapeutic (60 mg) and sub-therapeutic (30 mg) doses and was likely aggravated further by polypharmaceutical interventions she was prescribed. In addition, the subject’s other conditions, such as sleep apnea, may also have contributed to her sleep disturbances and general physical and psychological health. While sleep disruption is a common side effect of SSRIs and SNRIs, our finding that Cymbalta appears to have exacerbated the subject’s condition, is important for personalized care of patients with nuanced conditions. The problems associated with Cymbalta may have been due to the extended release formulation of the drug. It is known that Cymbalta is metabolized by CYP2D6, which has been recently shown to undergo a metabolizer phenotype conversion that cannot be assessed by genetic testing50. Drug-induced and particularly co-medication-induced phenoconversion is an increasing problem for personalized medicine51. Additionally, temazepam is not a short-acting benzodiazepine drug and can cause hangover effects in the course of a night that could contribute to the phase-delay our subject experienced. In fact, both temazepam and another highly used sleep aid, Ambien, were recently found to be associated with increased morbidity and mortality52. Despite the fact our subject was co-morbid for a number of circadian disruptors, her sleep architecture normalized when all drugs were removed. In addition, drug removal unmasked substantial sleep apnea, manifesting mainly during an NREM sleep component. The temazepam-Cymbalta combination appears to have induced a removal of deep sleep that actually mimics the shallow sleep architecture seen in depressed patients53. Antidepressants are often touted as able to restore deep sleep and delay REM sleep in depression53. However, for the subject of focus here (and we suspect many others), the major destruction of her deep sleep occurred when a sleep aid was added to counteract the over-stimulation of the antidepressant.\n\nA number of studies have shown that antidepressants can exacerbate symptoms associated with depression30–37. Further, we found that our subject suffered from obstructive sleep apnea (OSA) and should probably never have been on sleep medication in the first place. After drug removal a substantial change in HRV was seen once OSA was relieved. In this light, it is difficult to know how much of the subject’s SAD symptomology was exacerbated by co-morbid obstructive sleep apnea. Symptom clusters of poor sleep, migraines, and fatigue should motivate a physician to perform a sleep study. In fact, both in menopausal women and in psychiatric practice where mood and sleep disorders can show bi-directional causation, ordering sleep studies for patients has become the recommended course54,55. However, this is not widely practiced at the primary care level. Ironically, during our study the subject was initially denied a sleep study and had to convince her HMO (using graphics from this study) before she was given an in-home sleep test for sleep apnea.\n\nMany disorders, such as SAD, dysthymia, depression, bipolar disorder, etc. require that regular eating, sleeping and exercise schedules are to be adhered to so as not to disrupt circadian rhythms and exacerbate symptoms. This is also why these disorders are also exacerbated by high rates of external stress, particularly in the winter. Newer approaches to endophenotyping individuals are underway and biomarkers that are not part of the normal clinical biochemistry panel need to be measured to track symptoms and their origins, such as melatonin and cortisol. For example, at 9 a.m., individuals with an ‘evening chronotype’ can still have significantly higher serum melatonin levels than individuals with a ‘morning chronotype’56 such that determining dim light melatonin onset (DLMO) is crucial57.\n\nThe introduction of the MDDscore for depression, based on 9 serum biomarkers (α1 antitrypsin, apolipoprotein C3, brain derived neurotropic factor, cortisol, epidermal growth factor, myeloperoxidase, prolactin, resistin, soluble tumor necrosis factor α receptor type II), is a start58. It does not, however, test for inflammatory cytokines (IL-1α, IL-1β, IL-6, TNFα) and C-reactive protein, which may be predictors of antidepressant response59, as inflammation is increasingly recognized as a factor in psychiatric disorders and may also be linked to phenoconversion of drug metabolizing enzymes60. Thus, rather than resulting to polypharmacy by default when considering treatment options for a patient, collecting periodic metabolic data could uncover the individual perturbations in key pathways, which could aid in identifying simple dietary or targeted treatments.\n\n\nLimitations\n\nThe drug withdrawal protocol for the subject discussed here ran from December to July. The days were getting longer across the time period (after winter solstice to after summer solstice) so changes in the subject’s responses to light and increased/decreased internal secretion of melatonin/serotonin could have had a beneficial influence on the direction of the changes in sleep parameters in parallel with drug removal. Alternatively, the hypersomnia expected in a SAD-susceptible individual during December-May could result in a more sound sleep (except for sleep latency issues expected from her phenotype/chronotype). However, we showed that the final (no-drug) sleep architecture in July 2013 was equivalent to that observed at the beginning of our sleep apnea intervention in December 2013. In the end, the subject demonstrated what is typical for SAD, normal sleep architecture, but tendency toward a delayed chronotype.\n\nDue to the free-living nature of our study, attempts to follow/collect standardized food, exercise and sleep/wake behavior were not maintained, although, attempts to phase-shift to earlier sleep/wake regimens were documented. There was no drug placebo, blinding or washouts between trials, but we were able to compare our subject’s status to her status at times when no drug was provided in a crossover setting. Abrupt changes in treatment may have contributed noise to the data, for example: 1) mouthguard changes led to jaw pain, hyper-salivation, etc. and 2) change from Cymbalta 60 mg to Cymbalta 30 mg caused hot-flashes and thus were factors impacting sleep initiation. The night of the Thanksgiving holiday (11-28-2013, during MG4x) exhibited increased HRV value outliers across all sleep stages, and so those data points were removed (see Supplementary Material). This phenomenon was probably due to increased rest and the intake of enriched food, highlighting the importance of dietary and lifestyle assessments in this disorder. For the most part, we collected enough data under each treatment studied (relative to drug or device on/off) to measure effects, including the capture of rebound and recovery effects, and the duration of our individual trial conditions were comparable to what is often seen in sleep literature.\n\n\nConclusions\n\nMany people suffering from circadian and sleep disturbances such as those found in SAD have very unique genetic determinants for their condition, different sets of sleep disturbance sequelae, secondary conditions, and nuanced lifestyles that make it hard to treat them exactly the same way. As a result, more focused attention on what intervention strategy makes the most sense to pursue is required. Such ‘personalized’ intervention strategies are not trivial to implement since they require an integrated, objective, and often-times completely empirical approach to identify and implement them. We describe our experience with, and the results of, a comprehensive investigation into the response of a single patient to designed manipulations of her sleep pharmacology. We find that the patient had underlying conditions (e.g., sleep apnea) that were confounded by the use of specific drugs to treat her SAD and that these drugs contributed to, or exacerbated, other issues in the subject’s life (e.g., alert time for work, attempts to make up for lack of quality sleep during the week on the weekends, etc.). Ultimately, our study and its results should set a precedent for patient-oriented, yet designed and objective, investigations into the impact of polypharmacy and general drug response in real-world settings.\n\n\nData availability\n\nF1000Research: Dataset 1. Drug dosage and sleep response data, 10.5256/f1000research.7694.d11201661\n\nF1000Research: Dataset 2. Mouthguard intervention, heart rate variability and sleep response data, 10.5256/f1000research.7694.d11201762\n\nF1000Research: Dataset 3. PAM-RL Periodic Leg Movement Rates, 10.5256/f1000research.7694.d11201863\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and/or clinical images was obtained from the patient.",
"appendix": "Author contributions\n\n\n\nVLM (\"the subject\") and NJS conceived the idea for the study. VLM designed and implemented experiments, collected and processed all device data, produced graphics and performed all final statistical analyses used in the manuscript, wrote and edited the manuscript. YW performed informatic integration of cross-platform data, initial descriptive, serial correlation and outlier statistical analyses and edited the manuscript. NJS directed all statistical analyses, wrote and the edited manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\n\n\nGrant information\n\nDrs. Magnuson and Schork are supported in part by the following NIH grants, U19 AG023122-09; R01 DA030976-05; R01 MH094483-03; R01 MH100351-02; R21 AG045789-01A1; UL1TR001442-01; U24AG051129-01; as well as grants from the Tanner Foundation.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to acknowledge the generous loan of demonstration (\"demo\") monitoring equipment and related software used in this study by Philips Respironics (Actiwatch Spectrum, Equivital, PAM-RL) and items donated by MD Revolution (Zeo Sleep Monitor, Fitbit, iPAD).\n\n\nSupplementary Material\n\nThe number of times per night subject was awake during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A-N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nThe number of hours (h) per night before subject achieved first REM sleep bout during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A-N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nThe percent time subject was awake during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A-N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nThe percent time subject was in REM sleep during 5-minute intervals detected by the Zeo Sleep Monitor throughout the entire study. Dosages of Cymbalta (CYM60 = 60 mg, CYM30 = 30 mg), temazepam (TEM30 = 30 mg, TEM15 = 15 mg) and melatonin (MEL3 = 3 mg, MEL6 = 6 mg) were varied according to combinations A-N (T1, T2, T3 are trial replicates), including no drug trials (L, N).\n\nAdjusted mean in milliseconds (ms), mean estimate, standard error (SE) and p-value (Pr > |t-value|). R2: R-squared.\n\nClinical evaluations suggested that the subject has a deviated septum, a small jaw with substantial retrognathia (overbite), and evidence of clenching and grinding her teeth during sleep. The subject was diagnosed with mild obstructive sleep apnea in August 2013 and a mandibular splint (mouth guard or MG) was fabricated as an intervention. The subject then wore the device as ordered by her physician. After an initial adjustment period, the subject was monitored while the splint was advanced to achieve relief of apnea symptoms (mainly snoring). The monitoring took place for 40 nights from 11-10-2013 through 12-19-2013. The unadjusted MG was denoted as MG0x for analysis purposes. Simply inserting a mouth guard creates vertical displacement of the mouth and jaw and also (by design) some horizontal displacement of the jaw. Subsequent adjustments of 4 turns and 6 turns of screws to advance the jaw were denoted as MG4x and MG6x, respectively. Four nights during the time the MG was considered in the analyses had missing data, three nights had no MG wear and 33 nights had the MG worn at settings 0x (13 days), 4x (11 days), and 6x (9 days).\n\nRespiratory sinus arrhythmia is a coupling of the heart rate to follow the breath rate (cardio-pulmonary coupling, CPC). Measures of CPC are usually collected during NREM (light + deep) sleep, which occurs mainly during the first half of the night. In order to accommodate this biological phenomenon and in an attempt to equalize the amount of the data generated by the 15-second Equivital heart rate collection rate, we used an “end-of-night” truncation of 6 am. The collection period used for data analysis was from first sleep bout to last sleep bout before 6 am. This resulted in a minimal change in sleep stage percentages. For example, the 0 mg Cymbalta mean estimates for percent wake (10.5 percent), light (35.4 percent), deep (22.3 percent) and REM (34.2 percent) stages from Table 5 average 15 percent, 33 percent, 23 percent and 30 percent, respectively, when nights are truncated to 6 am. The truncated data shows decreases in percent light and REM sleep which are more prevalent during end-of-night sleep.\n\nHeart rate variability (HRV) is decreased during sleep apnea as the breath is obstructed. An increase in HRV should be seen when obstructive sleep apnea is treated with mandibular advancement. A commonly used measure of HRV is the standard deviation of the R-R (beat-to-beat) interval in milliseconds (ms), also called SDNN. The histograms of R-R interval average (RR-AVG) and the R-R interval standard deviation (RR-STDEV) when grouped by MG adjustments appeared to be roughly normally distributed. One night appeared to be a significant outlier across HRV observations (Thanksgiving night, 11-28-2013, MG setting = 4x) and was removed from all HRV analyses. Excluding this night had a substantial impact on improving model R2 values, but did not change the overall relationship between HRV and MG settings. We analyzed HRV (RR-STDEV) across the entire night and because HRV might be expected to differ among the different sleep stages, we also analyzed the impact of the MG by sleep stage. Univariate regression analyses were performed as previously described (see Methods) and assessments showed that all linear regression assumption requirements were satisfied. Table S1 shows a significant increase in total nightly HRV at the MG0x, MG4x and MG6x advanced settings. Regression coefficients are expressed in the original unit of ms for discussion and treatment effects have been adjusted relative to the intercept. The estimate of the y-intercept (μ0) for the model with HRV as the dependent variable suggests that the subject’s HRV was approximately 40ms while wearing no mouth guard. The estimated coefficients for the MG settings as independent variables in the model provide the effect on HRV of the mouth guard. The subject’s mean HRV increased to approximately 47ms (7.16 (SE: 2.83), p = 0.0168), 47ms (6.76 (SE: 2.91), p = 0.0270) and 54ms (13.87 (SE: 2.95), p = 5.0×10-5) while wearing the mouth guard at the MG0x, MG4x and MG6x settings, respectively, for a total nightly increase of 13.87ms in HRV. The breakdown by sleep stage suggested that compared to no mouth guard, MG6x increased the subject’s mean HRV from approximately 54ms to 66ms (12.22 (SE: 3.12), p = 0.0004) in wake; from approximately 45ms to 55ms (10.20 (SE: 3.30), p = 0.0040) in light and from approximately 35ms to 44ms (9.20 (SE: 2.68), p = 0.0017) in deep sleep stages. Interestingly, there was no effect on HRV during REM sleep. This confirms that the effects of MG use we observed are confined to the stages of sleep (NREM) that we felt were expected. None of the models were improved by adding in either night of the study or day of the week.\n\nAdjusted mean in percent, transformed mean estimatebc, standard error (SE) and p-value (Pr > |t-value|). R2: R-squared; bc: Box-Cox transformed variable raised to exponent given in final model. Back-transformation to original units was performed (after adjustments relative to intercept) by taking the nth (exponent) root of estimate.\n\nWe also determined the impact that treating sleep apnea had on our subject’s sleep quality. Table S2 shows the effect of MG setting on sleep stages. Because of the short time period tested at each MG setting and the non-normality of the percent sleep data, the Box-Cox procedure and transformation was performed, and models refit as above. As above, for discussion, the mean estimates presented in Table S2 are derived from the regression coefficients which have been adjusted and back-transformed to give the original unit of percent.\n\nThe estimate of the y-intercept (μ0) for the model with wake as the dependent variable and estimates of the coefficients for MG settings as independent variables suggest that the subject’s mean percent wake steadily decreased from approximately 15.6 percent (0.52 (SE: 0.04), p < 1.9×10-14) to 7.4 percent (-0.12 (SE: 0.04), p = 0.0021), 4.6 percent (-0.18 (SE: 0.04), p=0.0002) and 2.7 percent (-0.24 (SE: 0.06), p = 0.0004) during the use of MG0x, MG4x and MG6x, respectively.\n\nThe estimate of the y-intercept (μ0) for the model with light sleep as the dependent variable and estimates of the coefficients for MG settings as independent variables suggest that the subject’s mean percent light sleep increased from approximately 33.3 percent (0.19 (SE: 0.01), p < 2×10-16) to 39.3 percent (0.05 (SE: 0.02), p = 0.0089) during the use of MG4x and 39 percent (0.05 (SE: 0.02), p = 0.0185) during the use of MG6x. Interestingly, MG use had no significant impact on percent deep or percent REM sleep.\n\nDespite the fact that the 0 mg Cymbalta pharmaceutical intervention portion of the study ended in July 2013, we calculated (above) that the subject was in wake approximately 15 percent of the time and in light sleep approximately 33 percent of the time when nights were truncated at 6 am, which matches the y-intercept estimates (μ0) for the no mouth guard state of the sleep apnea study performed from November-December 2013 (Table S2, 15.6 and 33.3 percent, respectively).\n\nFinally, although we were unable to model percent deep and REM sleep during MG wear, the calculated mean percent deep sleep was 23 percent and the calculated mean percent REM sleep was 29 percent for the 40-night period, which also compares well with the averages calculated from the truncated data (23 percent and 30 percent, respectively). These data indicate a very close agreement in measurements despite an interval of several months and the change in season between the two studies and provides a certain comfort level for the purposes of comparing results from the two interventions.\n\n\nReferences\n\nRoecklein KA, Rohan KJ: Seasonal affective disorder: an overview and update. Psychiatry (Edgmont). 2005; 2(1): 20–6. PubMed Abstract | Free Full Text\n\nDanilenko KV, Levitan RD: Seasonal affective disorder. Handb Clin Neurol. 2012; 106: 279–89. PubMed Abstract | Publisher Full Text\n\nChen WF, Majercak J, Edery I: Clock-gated photic stimulation of timeless expression at cold temperatures and seasonal adaptation in Drosophila. J Biol Rhythms. 2006; 21(4): 256–71. PubMed Abstract | Publisher Full Text\n\nAguilar-Arnal L, Sassone-Corsi P: The circadian epigenome: how metabolism talks to chromatin remodeling. Curr Opin Cell Biol. 2013; 25(2): 170–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu Y, Hu W, Murakawa Y, et al.: Cold-induced RNA-binding proteins regulate circadian gene expression by controlling alternative polyadenylation. Sci Rep. 2013; 3: 2054. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCiuzan O, Hancock J, Pamfil D, et al.: The evolutionarily conserved multifunctional glycine-rich RNA-binding proteins play key roles in development and stress adaptation. Physiol Plant. 2015; 153(1): 1–11. PubMed Abstract | Publisher Full Text\n\nWehr TA, Duncan WC Jr, Sher L, et al.: A circadian signal of change of season in patients with seasonal affective disorder. Arch Gen Psychiatry. 2001; 58(12): 1108–14. PubMed Abstract | Publisher Full Text\n\nLevitan RD: The chronobiology and neurobiology of winter seasonal affective disorder. Dialogues Clin Neurosci. 2007; 9(3): 315–24. PubMed Abstract | Free Full Text\n\nRoecklein KA, Wong PM, Miller MA, et al.: Melanopsin, photosensitive ganglion cells, and seasonal affective disorder. Neurosci Biobehav Rev. 2013; 37(3): 229–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoecklein K, Wong P, Ernecoff N, et al.: The post illumination pupil response is reduced in seasonal affective disorder. Psychiatry Res. 2013; 210(1): 150–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcMahon B, Andersen SB, Madsen MK, et al.: P.1.i.037 Patients with seasonal affective disorder show seasonal fluctuations in their cerebral serotonin transporter binding. Eur Neuropsychopharm. 2014; 24(Suppl 2): S319. Publisher Full Text\n\nGoldman BD: Pattern of melatonin secretion mediates transfer of photoperiod information from mother to fetus in mammals. Sci STKE. 2003; 2003(192): PE29. PubMed Abstract | Publisher Full Text\n\nFoster RG, Roenneberg T: Human Responses to the Geophysical Daily, Annual and Lunar Cycles. Curr Biol. 2008; 18(17): R784–R94. PubMed Abstract | Publisher Full Text\n\nCiarleglio CM, Axley JC, Strauss BR, et al.: Perinatal photoperiod imprints the circadian clock. Nat Neurosci. 2011; 14(1): 25–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJackson CR, Capozzi M, Dai H, et al.: Circadian perinatal photoperiod has enduring effects on retinal dopamine and visual function. J Neurosci. 2014; 34(13): 4627–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreen NH, Jackson CR, Iwamoto H, et al.: Photoperiod programs dorsal raphe serotonergic neurons and affective behaviors. Curr Biol. 2015; 25(10): 1389–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMårtensson B, Pettersson A, Berglund L, et al.: Bright white light therapy in depression: A critical review of the evidence. J Affect Disord. 2015; 182: 1–7. PubMed Abstract | Publisher Full Text\n\nFisher PM, Madsen MK, Mc Mahon B, et al.: Three-week bright-light intervention has dose-related effects on threat-related corticolimbic reactivity and functional coupling. Biol Psychiatry. 2014; 76(4): 332–9. PubMed Abstract | Publisher Full Text\n\nLewy AJ, Emens JS, Songer JB, et al.: Winter Depression: Integrating mood, circadian rhythms, and the sleep/wake and light/dark cycles into a bio-psycho-social-environmental model. Sleep Med Clin. 2009; 4(2): 285–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalters JF, Hampton SM, Ferns GA, et al.: Effect of menopause on melatonin and alertness rhythms investigated in constant routine conditions. Chronobiol Int. 2005; 22(5): 859–72. PubMed Abstract | Publisher Full Text\n\nEichling PS, Sahni J: Menopause related sleep disorders. J Clin Sleep Med. 2005; 1(3): 291–300. PubMed Abstract\n\nCohen LS, Soares CN, Vitonis AF, et al.: Risk for new onset of depression during the menopausal transition: the Harvard study of moods and cycles. Arch Gen Psychiatry. 2006; 63(4): 385–90. PubMed Abstract | Publisher Full Text\n\nBixler EO, Vgontzas AN, Lin HM, et al.: Prevalence of sleep-disordered breathing in women: effects of gender. Am J Respir Crit Care Med. 2001; 163(3 Pt 1): 608–13. PubMed Abstract | Publisher Full Text\n\nDancey DR, Hanly PJ, Soong C, et al.: Impact of menopause on the prevalence and severity of sleep apnea. Chest. 2001; 120(1): 151–5. PubMed Abstract | Publisher Full Text\n\nDursunoglu N, Dursunoglu D: Do we neglect women with sleep apnea? Maturitas. 2007; 56(3): 332–4. PubMed Abstract | Publisher Full Text\n\nKapsimalis F, Kryger M: Sleep breathing disorders in the U.S. female population. J Womens Health (Larchmt). 2009; 18(8): 1211–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNowakowski S, Meers J, Heimbach E: Sleep and Women’s Health. Sleep Med Res. 2013; 4(1): 1–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMehta PK, Wei J, Wenger NK: Ischemic heart disease in women: a focus on risk factors. Trends Cardiovasc Med. 2015; 25(2): 140–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPark KE, Pepine CJ: Assessing cardiovascular risk in women: looking beyond traditional risk factors. Trends Cardiovasc Med. 2015; 25(2): 152–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArgyropoulos SV, Wilson SJ: Sleep disturbances in depression and the effects of antidepressants. Int Rev Psychiatry. 2005; 17(4): 237–45. PubMed Abstract | Publisher Full Text\n\nWilson S, Argyropoulos S: Antidepressants and sleep: a qualitative review of the literature. Drugs. 2005; 65(7): 927–47. PubMed Abstract | Publisher Full Text\n\nLam RW: Sleep disturbances and depression: a challenge for antidepressants. Int Clin Psychopharmacol. 2006; 21(Suppl 1): S25–9. PubMed Abstract | Publisher Full Text\n\nDeMartinis NA, Winokur A: Effects of psychiatric medications on sleep and sleep disorders. CNS Neurol Disord Drug Targets. 2007; 6(1): 17–29. PubMed Abstract | Publisher Full Text\n\nThase ME, Murck H, Post A: Clinical relevance of disturbances of sleep and vigilance in major depressive disorder: a review. Prim Care Companion J Clin Psychiatry. 2010; 12(6): pii: PCC.08m00676. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWichniak A, Wierzbicka A, Jernajczyk W: Sleep and antidepressant treatment. Curr Pharm Des. 2012; 18(36): 5802–17. PubMed Abstract | Publisher Full Text\n\nYang C, White DP, Winkelman JW: Antidepressants and periodic leg movements of sleep. Biol Psychiatry. 2005; 58(6): 510–4. PubMed Abstract | Publisher Full Text\n\nKierlin L, Littner MR: Parasomnias and antidepressant therapy: a review of the literature. Front Psychiatry. 2011; 2: 71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAsnis GM, Chakraburtty A, DuBoff EA, et al.: Zolpidem for persistent insomnia in SSRI-treated depressed patients. J Clin Psychiatry. 1999; 60(10): 668–76. PubMed Abstract | Publisher Full Text\n\nLondborg PD, Smith WT, Glaudin V, et al.: Short-term cotherapy with clonazepam and fluoxetine: anxiety, sleep disturbance and core symptoms of depression. J Affect Disord. 2000; 61(1–2): 73–9. PubMed Abstract | Publisher Full Text\n\nSmith WT, Londborg PD, Glaudin V, et al.: Is extended clonazepam cotherapy of fluoxetine effective for outpatients with major depression? J Affect Disord. 2002; 70(3): 251–9. PubMed Abstract | Publisher Full Text\n\nLillie EO, Patay B, Diamant J, et al.: The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med. 2011; 8(2): 161–173. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchork NJ: Personalized medicine: Time for one-person trials. Nature. 2015; 520(7549): 609–11. PubMed Abstract | Publisher Full Text\n\nShambroom JR, Fábregas SE, Johnstone J: Validation of an automated wireless system to monitor sleep in healthy adults. J Sleep Res. 2012; 21(2): 221–30. PubMed Abstract | Publisher Full Text\n\nGriessenberger H, Heib DP, Kunz AB, et al.: Assessment of a wireless headband for automatic sleep scoring. Sleep Breath. 2013; 17(2): 747–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRhee MK, Lee HJ, Rex KM, et al.: Evaluation of two circadian rhythm questionnaires for screening for the delayed sleep phase disorder. Psychiatry Investig. 2012; 9(3): 236–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchramm PJ, Thomas R, Feige B, et al.: Quantitative measurement of sleep quality using cardiopulmonary coupling analysis: a retrospective comparison of individuals with and without primary insomnia. Sleep Breath. 2013; 17(2): 713–21. PubMed Abstract | Publisher Full Text\n\nKufoy E, Palma JA, Lopez J, et al.: Changes in the heart rate variability in patients with obstructive sleep apnea and its response to acute CPAP treatment. PLoS One. 2012; 7(3): e33769. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPalma JA, Urrestarazu E, Lopez-Azcarate J, et al.: Increased sympathetic and decreased parasympathetic cardiac tone in patients with sleep related alveolar hypoventilation. Sleep. 2013; 36(6): 933–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInstitute of Medicine (US) Committee on Sleep Medicine and Research, Colten HR, Altevogt BM: Sleep Disorders and Sleep Deprivation: An Unmet Public Health Problem. Washington, DC: Institute of Medicine: National Academies Press. 2006; xviii, 404. PubMed Abstract\n\nPreskorn SH, Kane CP, Lobello K, et al.: Cytochrome P450 2D6 phenoconversion is common in patients being treated for depression: implications for personalized medicine. J Clin Psychiatry. 2013; 74(6): 614–21. PubMed Abstract | Publisher Full Text\n\nShah RR, Smith RL: Addressing phenoconversion: the Achilles’ heel of personalized medicine. Br J Clin Pharmacol. 2015; 79(2): 222–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKripke DF, Langer RD, Kline LE: Hypnotics’ association with mortality or cancer: a matched cohort study. BMJ Open. 2012; 2(1): e000850. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSlaughter JR: Sleep and depression. Mo Med. 2006; 103(5): 526–8. PubMed Abstract\n\nBruyneel M: Sleep disturbances in menopausal women: Aetiology and practical aspects. Maturitas. 2015. 81(3): 406–9. PubMed Abstract | Publisher Full Text\n\nKrystal AD: Psychiatric disorders and sleep. Neurol Clin. 2012; 30(4): 1389–413. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorera-Fumero AL, Abreu-González P, Henry-Benitez M, et al.: Chronotype as modulator of morning serum melatonin levels. Actas Esp Psiquiatr. 2013; 41(3): 149–53. PubMed Abstract\n\nKeijzer H, Smits MG, Duffy JF, et al.: Why the dim light melatonin onset (DLMO) should be measured before treatment of patients with circadian rhythm sleep disorders. Sleep Med Rev. 2014; 18(4): 333–9. PubMed Abstract | Publisher Full Text\n\nBilello JA, Thurmond LM, Smith KM, et al.: MDDScore: confirmation of a blood test to aid in the diagnosis of major depressive disorder. J Clin Psychiatry. 2015; 76(2): e199–e206. PubMed Abstract | Publisher Full Text\n\nHashimoto K: Inflammatory biomarkers as differential predictors of antidepressant response. Int J Mol Sci. 2015; 16(4): 7796–801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShah RR, Smith RL: Inflammation-induced phenoconversion of polymorphic drug metabolizing enzymes: hypothesis with implications for personalized medicine. Drug Metab Dispos. 2015; 43(3): 400–10. PubMed Abstract | Publisher Full Text\n\nMagnuson V, Wang Y, Schork N: Dataset 1 in: Normalizing sleep quality disturbed by psychiatric polypharmacy and sleep apnea: a comprehensive patient-centered N-of-1 study. F1000Research. 2016. Data Source\n\nMagnuson V, Wang Y, Schork N: Dataset 2 in: Normalizing sleep quality disturbed by psychiatric polypharmacy and sleep apnea: a comprehensive patient-centered N-of-1 study. F1000Research. 2016. Data Source\n\nMagnuson V, Wang Y, Schork N: Dataset 3 in: Normalizing sleep quality disturbed by psychiatric polypharmacy and sleep apnea: a comprehensive patient-centered N-of-1 study. F1000Research. 2016. Data Source"
}
|
[
{
"id": "12973",
"date": "11 May 2016",
"name": "Wilson D. Pace",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript purports to report on an N-of-1 trial to select therapy for a patient with seasonal affective disorder, sleep apnea and polypharmacy. The outcome of the activities was improved sleep quality. The condition, the treatments to be considered and the outcomes all appear to be excellent concepts to submit to N-of-1 trials. The interventional approaches are described in detail. The medication trial is the closest this process comes to a true N-of-1 trial. This said, the entire manuscript appears to be an early draft that requires extensive rework.This appears to be a complex case study with some quasi-experimental components of the various intervention approaches. The current version of the manuscript mixes the background with the methods section, the methods section with the results section and the discussion covers interventions outside of the medication trials or the manuscript in general. Again, the medication trial is the only part of the process that approaches an N-of-1 trial and should be the focus of the manuscript.The current manuscript is very long and difficult to follow. The current draft is just under 10,000 words for the primary paper, excluding the abstract, supplemental material and references. This manuscript would be much easier to follow and comprehend if cut to approximately 3500-4000 words, which is already long for a medical article. This will require extensive editing and decisions about what to include and what to exclude. This reviewer cannot provide full editing guidance but the authors need to consult with others that can help craft future versions.Areas that need to be addressed:The extensive description of the patient can be markedly reduced. Further this component of the manuscript should be in the Methods section as it is essentially a description of the study population. The description should focus on the state of affairs just prior to initiating the medication trial. The remaining background is essentially irrelevant to this case study. The closest part of the therapeutic process that approaches an N-of-1 trial is the medication component. This reviewer recommends focusing on this component of the work if the paper is to be retained as an N-of-1 trail. With this change all the interventions that are discussed prior to or outside of this set of interventions can be dropped and included as the state of the study participant at the start of the trial. The extensive discussion of measurement activities needs to significantly cut and measurement approaches referenced from other literature. The discussion of the measurement approaches is also included in the results section as well as the methods section. Some of the methods section related to the sleep apnea treatment intervention appear to be results in the current draft. This can be solved by dropping the extensive discussion of the sleep apnea diagnosis and intervention entirely as it was not an N-of-1 trial in any sense.The methods should discuss the N-of-1 approach that was used. The decision to not blind medications should be justified. The cross over pattern selection should discussed. It appears the number of crossovers for each treatment option is limited. This is the primary reason this manuscript appears to be more a case study than a true experimental approach. The number of crossovers should be justified, especially for those medications options that were only studied one time. It appears that the medications were studied primarily in a series of reductions in dosages until the final dosages where there was repeat testing in a back and forth pattern. N-of-1 crossover patterns should be randomized, thus the testing pattern needs to be justified. Given the high variability in sleep quality from night to night the decision to use a limited number of cross-overs seems even more troublesome. The reason for not considering a washout period between treatments needs to be justified.The discussion of serum markers for major depression disorder is not related to the study methods or results and should be removed. Limitations related to the limited number of crossovers is not discussed. The interference of the mouth guard intervention with the medication trial further complicates the low number of drug crossovers.This manuscript requires major editing and rewriting prior to being reconsidered.",
"responses": [
{
"c_id": "2119",
"date": "05 Sep 2016",
"name": "Victoria Magnuson",
"role": "Author Response",
"response": "We would like to thank Dr. Wilson Pace for his insightful review of our manuscript and his many constructive suggestions for improvement. We have amended our manuscript in response and believe it is a much better article as a result."
}
]
}
] | 1
|
https://f1000research.com/articles/5-132
|
https://f1000research.com/articles/5-2226/v1
|
05 Sep 16
|
{
"type": "Opinion Article",
"title": "Clonal selection versus clonal cooperation: the integrated perception of immune objects",
"authors": [
"Serge Nataf"
],
"abstract": "Analogies between the immune and nervous systems were first envisioned by the immunologist Niels Jerne who introduced the concepts of antigen \"recognition\" and immune \"memory\". However, since then, it appears that only the cognitive immunology paradigm proposed by Irun Cohen, attempted to further theorize the immune system functions through the prism of neurosciences. The present paper is aimed at revisiting this analogy-based reasoning. In particular, a parallel is drawn between the brain pathways of visual perception and the processes allowing the global perception of an \"immune object\". Thus, in the visual system, distinct features of a visual object (shape, color, motion) are perceived separately by distinct neuronal populations during a primary perception task. The output signals generated during this first step instruct then an integrated perception task performed by other neuronal networks. Such a higher order perception step is by essence a cooperative task that is mandatory for the global perception of visual objects. Based on a re-interpretation of recent experimental data, it is suggested that similar general principles drive the integrated perception of immune objects in secondary lymphoid organs (SLOs). In this scheme, the four main categories of signals characterizing an immune object (antigenic, contextual, temporal and localization signals) are first perceived separately by distinct networks of immunocompetent cells. Then, in a multitude of SLO niches, the output signals generated during this primary perception step are integrated by TH-cells at the single cell level. This process eventually generates a multitude of T-cell and B-cell clones that perform, at the scale of SLOs, an integrated perception of immune objects. Overall, this new framework proposes that integrated immune perception and, consequently, integrated immune responses, rely essentially on clonal cooperation rather than clonal selection.",
"keywords": [
"theoretical immunology",
"neuroimmunology",
"sensory perception",
"immunity",
"brain",
"T-cells"
],
"content": "Introduction\n\nEvolution has endowed the human species with the most sophisticated immune and nervous systems. Maintenance of our internal homeostasis and adaptation to our external environment rely essentially on the ability of both systems to sense, memorize and react to a large variety of input signals. These crucial functions are supported by a common organizational grounding base consisting in complex networks of specialized cells that communicate in specific anatomical sites. Similarities between the immune and nervous systems were first highlighted by the immunologist Niels Jerne who introduced the terms “recognition”, “memory” and “learning” in the immunological vocabulary1,2. However, since then, such an analogy-based reasoning was mostly used to demonstrate reminiscent molecular mechanisms between immune and neuronal synapses3,4. Only few works attempted to further theorize the immune system functions through the prism of neurosciences. In this framework, the cognitive immunology paradigm proposed by Irun Cohen5,6 is undeniably a key contribution that notably led to the concept of physiological auto-immunity7,8. In particular, Irun Cohen proposed that naturally occurring auto-antibodies provide an indispensable immune system's representation of our body, the immunological homunculus5, which resembles its neural counterparts, the somatosensory homunculus. Thereafter, other works similarly apprehended immunity as a cognitive process and brought about the emergence of computational immunology9,10. Nevertheless, the line of thought initiated by Jerne appears not to have been nourished by the major conceptual and experimental advances that cognitive neurosciences provided in the last two decades. This context offers a unique opportunity to revisit and explore analogies between the nervous and immune systems in the light of such discoveries. The recently formulated concept of a sensory immune system11 falls into this re-thinking strategy.\n\nDeveloping further the concept of “perceptive immunity” requires beforehand to provide a basic description of the main mechanisms allowing our brain to perceive sensory inputs. Let us choose the example of visual perception. When considering the perception of a given visual object, different categories of input signals that relate with the shape, color, and motion of this object are captured and integrated independently by distinct neuronal populations. These specialized neuronal networks reside in the so-called primary visual cortex, in the superficial neuronal layers of the brain occipital lobe)12,13. Importantly, such a primary perception induces the generation of output primary signals (electro-chemical by nature) that converge toward neurons localized in the so-called visual association areas also named higher-order visual areas14,15 (Figure 1). There, these specialized neuronal populations capture and integrate varied combinations of output primary signals to perform an integrated perception of visual objects. Two major pathways allowing the convergence of output primary signals toward higher-order areas are well characterized: i) the “What” pathway targets associative areas in the temporal cortex and is essential to the recognition and memorization of visual objects14, ii) the “Where” pathway targets associative areas of the parietal cortex and supports the perception of precise localization14. Eventually, the interconnections between high-order visual areas allows a fully-integrated perception that takes into account the nature, localization, context and time-related features of a visual object14.\n\nThe signals related to the shape, color and motion of visual objects are integrated by specialized brain neuronal populations residing in the primary visual cortex i.e the superficial neuronal layers of the brain occipital lobe. Output signals are then generated that instruct other cortical areas for higher order integration tasks. The “What” pathway connects the primary visual cortex to areas of the temporal cortex that are essential to the recognition and memorization of objects and forms. The “Where” pathway connects the primary visual cortex to areas of the parietal cortex that support perception of precise localization. The interconnections between higher-order visual areas (dashed arrows) as well as other brain areas not highlighted here, allows a fully-integrated perception that takes into account the “What”, “Where”, “How” and “When” features of a visual object.\n\nThus, visual perception requires not only a specialization of cells depending on their ability to perform primary vs integrated perception tasks but also tight cooperation between neuronal networks. Primary perception allows distinct features of a visual object (shape, color, motion) to be perceived separately12,13. Integrated perception allows a visual object to be perceived as a whole via the integration of distinct categories of primary signals14,15.\n\nIt also important to underscore that the operability of any neuronal network, would it be involved or not in sensory perception, depends on non-neuronal cells that locate in close vicinity to neurons. Astrocytes exert a tight control of interneuronal synaptic transmission16,17 and microglia, the resident macrophages of the brain, proceed to a selective trimming of functionally irrelevant or supernumerary synapses18,19. In addition, the blood flow in small arteries and capillaries of the brain is exquisitely tuned by a mechanism of neurovascular coupling that finely adjusts the supply of blood-derived oxygen and glucose to the needs of neuronal networks20,21.\n\nPostulating the existence of analogies between the immune and visual systems implies first that the counterpart of visual objects are immune objects. If so, immune objects cannot be simply reduced to an antigen +/- danger signals. Indeed, in accordance with the principles of immunogenicity previously enunciated by Rolf Zinkernagel22, one may propose that an immune object (IO) is defined by the association of at least 4 categories of signals: antigenic, contextual, temporal and localization signals. Establishing a parallel between visual and immune perception also implies that the perception of an IO relies first on primary perception tasks followed by an integrated perception step. It is proposed that distinct IO-related features (antigen, context, localization, time-related signals) are perceived separately by distinct networks of immunocompetent cells in a myriad of SLO niches. Then, output signals generated by such a primary perception step converge toward T-cells which, at the scale of SLOs, perform and orchestrate the integrated perception of IOs.\n\nAs previously proposed11, SLOs are likely to be the main anatomical sites where the primary and integrated perception of IOs take place. Below is an attempt to categorize immunocompetent cells according to their functions in the primary perception of IOs and the ensuing generation of primary output signals (Figure 2).\n\nThe antigenic, temporo-contextual and localization signals that characterize an immune object are perceived separately by specific populations of cells that perform a primary perception task. The output signals generated by these cells will then instruct an integrated perception step essentially performed by TH-cells. The primary perception of antigenic signals is performed by DCs, macrophages, B-cells and any APC that may reside or migrate in lymph nodes. The primary perception of contextual and temporal signals including notably DAMP, PAMP and cytokines are performed by a large range of immune cells or non-immune cells that reside in SLOs. In addition, immune cells that target SLOs in a context- and time-dependent fashion also perform a primary perception of temporo-contextual signals. These cells include notably Treg and Breg cells irrespective of their target antigens as well as, to some extent, naive B- or T-cells. Finally, the primary perception of localization signals is essentially performed by APCs that derive from the immune object microenvironment.\n\n1) The primary perception of antigenic signals is essentially performed by antigen-presenting cells (APCs) would they belong or not to the dendritic cells (DC) lineage23. While tissue-resident DCs are the first line cells exerting such a function, a flurry of APCs that reside or migrate in SLOs also participate to the primary perception of antigens. Depending on the intrinsic properties harbored by APCs with regard to antigen processing and co-stimulation, the primary perception of antigens will result in the presentation of distinct epitopes and the expression of varied combinations of accessory molecules23.\n\n2) Contextual signals are highly diverse in nature and may combine in many different ways under physiological or pathological conditions (development, ageing, trauma, degeneration, infection, cancer…etc.). Danger-associated molecular patterns (DAMP), pathogen-associated molecular patterns (PAMP) and cytokines, which form the great majority of contextual danger signals, bind receptors harboring a large expression pattern in SLOs. The primary perception of contextual signals is thus likely to involve not only immune cells but also SLO-residing endothelial cells and stromal cells24–26. Output primary signals consist in a larger array of cell surface and soluble factors that instruct the integrated perception step. Moreover, a variety of immune cells that target SLOs in a context-dependent fashion participate to the primary perception of contextual signals and provide part of the primary response to such signals. These include notably Treg and Breg cells irrespective of their antigen specificity27,28, NK cells29, polymorphonuclear cells30, monocytes31,32, innate lymphoid cells33,34 as well as naive T or B lymphocytes35,36. Overall, the primary perception of contextual signals is a cooperative task performed by a large array of cell types that reside in SLOs or migrate toward SLOs. These cells generate of whole of soluble or membranous output signals that instruct the integrated perception of IOs.\n\n3) The primary perception of localization signals is essentially performed by dendritic cells and macrophages that are drained from the IO's tissue environment37–40. The output signals generated by these tissue-derived APCs will imprint the homing properties of T-cells38,39,41 and orientate in a tissue-specific manner the polarization of TH cells (notably toward TFH cells)42. Interestingly, recent findings indicate that stromal cells also play a major role in the primary perception of localization signals43,44.\n\n4) Temporal cues are provided by the duration of the antigenic, contextual and/or localization signals. Sudden vs chronically-installed IOs are not equally seen by the immune system according to the discontinuity theory45. However, one may also consider that the distinct patterns of tissue-derived cytokines characterizing acute vs subacute vs chronic inflammation provide crucial time-related inputs. In any case, as proposed above for contextual signals, the primary perception of temporal signals is a cooperative task performed by a large array of cell types in SLOs.\n\nThe highly complex spatial organization of SLOs, essentially determined by stromal cells24,46–48, is currently viewed as a means to tightly control the movements of cells and fluids in SLOs49,50. Such a stromal scaffold formed by endothelial cells and fibroblastic reticular cells also provides a histologic support to a number of niches that exhibit distinct microenvironments. For a given IO, there is indeed a myriad of APCs (subsets of DCs, macrophages, B-cells, other APCs) that interact with naïve or central memory T-cells in specific niches localized, for lymph nodes, in the paracortical51, subcapsular52,53 or medullar zone54,55. Such niches are formed by partially overlapping yet distinct compositions of immunocompetent cells that proceed to the primary perception of antigenic, contextual, localization and temporal signals. It can be proposed that in each of these niches, immunocompetent cells having proceeded to the primary perception of an IO generates output signals that converge toward T-cells bearing cognate TCRs. By this mean, T-cells (TH cells or T cytotoxic CD8 T-cells activated via cross-presentation) capture and integrate at the single cell level a whole of signals that relate with the antigenic, contextual, localization and temporal features of an IO.\n\nThe multitude of niche-specific integration processes that are performed at the single cell level in SLOs is likely to generate a variety of T-cell subpopulations harboring distinct functional behaviors (TH1, TH2, TH17, TFH, Treg…) and recognizing distinct immunodominant epitopes (Figure 3 and Figure 4). Supporting this view, two recent important studies definitively demonstrated that: i) CD4 T cells primed in vivo by pathogens or vaccines are highly heterogeneous with regard to TCRs and TH profiles56, ii) germinal center reactions in response to complex antigens generate a highly diverse B-cell population in terms of BCR affinity57. Thus, at the scale of a SLO the integrated perception of an IO relies on a multitude of antigen-specific T-cell and B-cell clones that provide a whole of \"angles of view\" from the same immune object. Whether or not such a diversity is progressively narrowed during the re-occurrence of an immune object (i.e during recall immune perception and recall immune response) would require investigations.\n\nEpitopes presented in the context of MHC molecules are the output signals resulting from the primary perception of antigens. In SLOs, multiple epitopes derived from an immune object are recognized by multiple epitope-specific T cells in distinct niches. These niches are formed by stromal cells (endothelial cells and fibroblastic reticular cells) and by partially overlapping combinations of immune cells performing the primary perception antigens, temporo-contextual signals and localization signals. Three examples of distinct SLO niches are depicted.\n\nThe interaction between APCs and epitope-specific T-cells occur in a multitude of niches that provide distinct combinations of output signals from primary perception tasks. In each niche, these output signals are integrated at the single cell level by epitope-specific T-cells. This step results in the generation of a large array of TH or CD8 T-cell subsets that possibly include anergic T-cells. Eventually, the integrated perception of IOs is performed at the scale of SLOs by a large variety of antigen-specific T-cell and B-cell clones. Only TH-cell clones are depicted in the diagram.\n\nIt is now recognized that a relatively high level of functional plasticity is maintained in transcriptionally committed TH cell subsets58–62 as well as in CD8 cytotoxic T-cells63 and B-cells64. While SLOs orchestrate an integrated immune perception of IOs, one may postulate that higher order integration steps may take place in the efferent lymphatic system. This process would rely on interclonal communications leading possibly to a functional \"cross-imprinting\" of TH cells (Figure 5).\n\nHigher order integration tasks may take place in the efferent lymphatic system. This process is proposed to rely on interclonal communications leading to a functional \"cross-imprinting\" of TH cells.\n\nAt the single cell level, a large number of immune cells may integrate antigenic, temporo-contextual and/or localization signals within or outside SLOs. Besides TH cells and CD8 T-cells, these include B-cells and plasma cells as well as NKT-cells and γδ T-cells. Moreover, recent findings indicate that innate myeloid or lymphoid cells may also integrate and memorize distinct categories of primary immune signals65. However, the immune perception theory sheds a new light on the obvious although frequently neglected statement that SLOs are indispensable to the generation of any integrated immune response and, in the context of perceptive immunity, any integrated immune perception. Indeed SLOs harbor a unique ability to: i) concentrate a large array of cells involved in primary integration tasks, ii) provide a multitude of niches for single cell integration processes.\n\nThe sensory nervous system allows perceiving as a whole the identity and nature of visual objects, their precise localization, visual context and time-related features (motion, memory traces). Visual perception and other facets of our sensory skills are functionally crucial in the orientation of decision making. Such an orientation may schematically follow three main axes: 1) neglect, 2) engage a neurocognitive activity (memorization, attention, thoughts, emotions…), 3) engage a motor activity (grasp, repel, approach, flee…) (Figure 6). Of note, visual perception is a dynamic process that not only orientates but continuously adjusts decision making. Thus, motor activity and visual perception are finely coupled via a whole of feedforward and feedback mechanisms allowing the execution of motor programs to be adjusted66. In a similar manner, the somatosensory perception of movements is essential to the control of motor activity67.\n\nDistinct primary perception tasks allow the shape, color and motion of a visual object to be perceived separately. Output signals generated from this primary perception step instruct an integrated perception allowing the “What”, “How”, “When” and “Where” of a visual object to be perceived as a whole. Such an integrated perception orientates decision making along three main axes: 1) neglect, 2) engage a neurocognitive activity (memorization, attention, thoughts, emotions…) or 3) engage a motor activity (grasp, repel, approach, flee…).\n\nSimilar to sensory neural perception, it may be suggested that the main function of immune perception is to orientate decision making toward the engagement of proper immune responses or immune programs. When bringing out the danger theory68,69, Polly Matzinger was the first to emphasize the importance of contextual inputs in the initiation of \"reject\" vs \"tolerate\" immune responses. Since then, the concept of “protective autoimmunity” enunciated by Michal Schwartz and Jonathan Kipnis70 stated that the recognition of tissue-specific auto-antigens allows the immune system to provide a tissue-specific support that is shaped by contextual signals71,72. A semantic adjustment to these major conceptual advances would consist in proposing that contextual inputs orientate decision making along 3 main axes: \"Reject\", \"Tolerate\" or \"Support\" i.e provide molecular and cellular instructing signals that maintain homeostasis73–75 or favor tissue repair76,77. A functional diagram of immune perception and decision making in the immune system could be then aligned with the model of visual perception and decision making in the nervous system (Figure 7). Along this line, it may be proposed that, similar to the visuomotor and sensorimotor feedback processes, effector immune cells that may be drained from tissues to SLOs deliver output signals reflecting the execution of immune programs. The efferent phase of any immune response would be thus constantly adjusted via feedback signals that are captured and integrated in SLOs.\n\nDistinct primary perception tasks allow the antigenic, temporo-contextual and localization signals characterizing an immune object to be perceived separately. Output signals generated from this primary perception step instruct an integrated perception allowing the \"What\", \"How\", \"When\" and \"Where\" of an immune object to be perceived as a whole. Such an integrated perception orientates decision making along three main axes: 1) tolerate, 2) reject, 3) support. i.e provide molecular and cellular instructing signals that maintain homeostasis or favor tissue repair.\n\nThe immune perception theory proposes that immunity is driven by several basic principles that are shared between the immune and nervous system. The first proposed principle is that immune cells are not only recognizing antigens +/- danger signals but are indeed perceiving immune objects that are formed by a whole of antigenic, contextual, temporal and localization signals. The second proposed principle is that immune signals are not only individually captured by immune cells but collectively integrated at the scale of SLOs. Such a cooperative functional organization holds relevance for the communications between innate and adaptive immune cells but also for the interactions between T-cell and B-cell clones that recognize a common immune object. The third basic principle is that immune perception is shaped by a number of parameters that are independent from the perceived immune objects11. These include notably the age, gender, metabolic status and gut microbiota composition of the host.\n\nOver the last decades, the research fields covered by immunology have considerably expanded along with the number of breakthrough discoveries relating with the immune system. As a consequence, capturing an up-to-date global image of the immune system functions has become an increasingly difficult task for education professionals and for students as well. In this regard, the theoretical framework proposed here may be essentially considered as a potentially valuable tool for the teaching of immunology. In addition, while neuroimmunology encompasses, for the most part, the study of neuroimmune interactions, the present work suggests that a larger partnership could be envisioned between neuroscientists and immunologists, on the realm of education. To face the challenge of intimately understanding complex systems such as the immune and nervous systems, a move toward an educational approximation of both disciplines is possibly of major importance to promote future cross-fertilizations of ideas and concepts.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work is supported by the Lyon-1 University and the Lyon University Hospital (Hospices Civils de Lyon), Lyon, France.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThanks to Nathalie Davoust-Nataf and Laurent Pays for their scientific and personal support.\n\n\nReferences\n\nJerne NK: Towards a network theory of the immune system. Ann Immunol (Paris). [Accessed October 19, 2015] 1974; 125C(1–2): 373–89. PubMed Abstract\n\nTauber A: The Biological Notion of Self and Non-self. [Accessed December 9, 2015] 2002. Reference Source\n\nChoudhuri K, Llodrá J, Roth EW, et al.: Polarized release of T-cell-receptor-enriched microvesicles at the immunological synapse. Nature. 2014; 507(7490): 118–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrakoui A, Bromley SK, Sumen C, et al.: The immunological synapse: a molecular machine controlling T cell activation. Science. [Accessed November 9, 2015] 1999; 285(5425): 221–7. PubMed Abstract | Publisher Full Text\n\nCohen IR: The cognitive paradigm and the immunological homunculus. Immunol Today. 1992; 13(12): 490–4. PubMed Abstract | Publisher Full Text\n\nCohen IR: The cognitive principle challenges clonal selection. Immunol Today. 1992; 13(11): 441–4. PubMed Abstract | Publisher Full Text\n\nMadi A, Bransburg-Zabary S, Maayan-Metzger A, et al.: Tumor-associated and disease-associated autoantibody repertoires in healthy colostrum and maternal and newborn cord sera. J Immunol. 2015; 194(11): 5272–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwartz M, Cohen IR: Autoimmunity can benefit self-maintenance. Immunol Today. [Accessed December 9, 2015] 2000; 21(6): 265–8. PubMed Abstract | Publisher Full Text\n\nBrownlee J: Cognition, Immunology, and the Cognitive Paradigm. 2007; 1–6. Reference Source\n\nCohen IR: Real and artificial immune systems: computing the state of the body. Nat Rev Immunol. 2007; 7(7): 569–74. PubMed Abstract | Publisher Full Text\n\nNataf S: The sensory immune system: a neural twist to the antigenic discontinuity theory. Nat Rev Immunol. 2014; 14(4): 280. PubMed Abstract | Publisher Full Text\n\nSincich LC, Horton JC: The circuitry of V1 and V2: integration of color, form, and motion. Annu Rev Neurosci. 2005; 28: 303–26. PubMed Abstract | Publisher Full Text\n\nLivingstone M, Hubel D: Segregation of form, color, movement, and depth: anatomy, physiology, and perception. Science. [Accessed November 19, 2015] 1988; 240(4853): 740–9. PubMed Abstract | Publisher Full Text\n\nde Haan EH, Cowey A: On the usefulness of ‘what’ and ‘where’ pathways in vision. Trends Cogn Sci. 2011; 15(10): 460–6. PubMed Abstract | Publisher Full Text\n\nKatzner S, Weigelt S: Visual cortical networks: of mice and men. Curr Opin Neurobiol. 2013; 23(2): 202–6. PubMed Abstract | Publisher Full Text\n\nAraque A, Carmignoto G, Haydon PG, et al.: Gliotransmitters travel in time and space. Neuron. 2014; 81(4): 728–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAraque A, Parpura V, Sanzgiri RP, et al.: Tripartite synapses: glia, the unacknowledged partner. Trends Neurosci. [Accessed November 29, 2015] 1999; 22(5): 208–15. PubMed Abstract | Publisher Full Text\n\nSchafer DP, Lehrman EK, Kautzman AG, et al.: Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron. 2012; 74(4): 691–705. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStephan AH, Barres BA, Stevens B: The complement system: an unexpected role in synaptic pruning during development and disease. Annu Rev Neurosci. 2012; 35: 369–89. PubMed Abstract | Publisher Full Text\n\nWälchli T, Wacker A, Frei K, et al.: Wiring the Vascular Network with Neural Cues: A CNS Perspective. Neuron. 2015; 87(2): 271–96. PubMed Abstract | Publisher Full Text\n\nAndreone BJ, Lacoste B, Gu C: Neuronal and vascular interactions. Annu Rev Neurosci. 2015; 38: 25–46. PubMed Abstract | Publisher Full Text\n\nZinkernagel RM: Localization dose and time of antigens determine immune reactivity. Semin Immunol. 2000; 12(3): 163–71; discussion 257–344. PubMed Abstract | Publisher Full Text\n\nRoche PA, Furuta K: The ins and outs of MHC class II-mediated antigen processing and presentation. Nat Rev Immunol. 2015; 15(4): 203–16. PubMed Abstract | Publisher Full Text\n\nMueller SN, Germain RN: Stromal cell contributions to the homeostasis and functionality of the immune system. Nat Rev Immunol. 2009; 9(9): 618–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCard CM, Yu SS, Swartz MA: Emerging roles of lymphatic endothelium in regulating adaptive immunity. J Clin Invest. 2014; 124(3): 943–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPegu A, Qin S, Fallert Junecko BA, et al.: Human lymphatic endothelial cells express multiple functional TLRs. J Immunol. [Accessed December 5, 2015] 2008; 180(5): 3399–405. PubMed Abstract | Publisher Full Text\n\nRosser EC, Oleinika K, Tonon S, et al.: Regulatory B cells are induced by gut microbiota-driven interleukin-1β and interleukin-6 production. Nat Med. 2014; 20(11): 1334–9. PubMed Abstract | Publisher Full Text\n\nSage PT, Paterson AM, Lovitch SB, et al.: The coinhibitory receptor CTLA-4 controls B cell responses by modulating T follicular helper, T follicular regulatory, and T regulatory cells. Immunity. 2014; 41(6): 1026–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaffont S, Seillet C, Ortaldo J, et al.: Natural killer cells recruited into lymph nodes inhibit alloreactive T-cell activation through perforin-mediated killing of donor allogeneic dendritic cells. Blood. 2008; 112(3): 661–71. PubMed Abstract | Publisher Full Text\n\nHampton HR, Bailey J, Tomura M, et al.: Microbe-dependent lymphatic migration of neutrophils modulates lymphocyte proliferation in lymph nodes. Nat Commun. 2015; 6: 7139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFörster R, Braun A, Worbs T: Lymph node homing of T cells and dendritic cells via afferent lymphatics. Trends Immunol. 2012; 33(6): 271–80. PubMed Abstract | Publisher Full Text\n\nJakubzick C, Gautier EL, Gibbings SL, et al.: Minimal differentiation of classical monocytes as they survey steady-state tissues and transport antigen to lymph nodes. Immunity. 2013; 39(3): 599–610. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMackley EC, Houston S, Marriott CL, et al.: CCR7-dependent trafficking of RORγ+ ILCs creates a unique microenvironment within mucosal draining lymph nodes. Nat Commun. 2015; 6: 5862. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang J, Hu S, Zhao L, et al.: Selective programming of CCR10+ innate lymphoid cells in skin-draining lymph nodes for cutaneous homeostatic regulation. Nat Immunol. 2016; 17(1): 48–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSckisel GD, Bouchlaka MN, Monjazeb AM, et al.: Out-of-Sequence Signal 3 Paralyzes Primary CD4+ T-Cell-Dependent Immunity. Immunity. 2015; 43(2): 240–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang CY, Vogt TK, Favre S, et al.: Trapping of naive lymphocytes triggers rapid growth and remodeling of the fibroblast network in reactive murine lymph nodes. Proc Natl Acad Sci U S A. 2014; 111(1): E109–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdele F, Molenaar R, Gütle D, et al.: Cutting edge: instructive role of peripheral tissue cells in the imprinting of T cell homing receptor patterns. J Immunol. [Accessed October 21, 2015] 2008; 181(6): 3745–9. PubMed Abstract | Publisher Full Text\n\nMora JR, Bono MR, Manjunath N, et al.: Selective imprinting of gut-homing T cells by Peyer’s patch dendritic cells. Nature. 2003; 424(6944): 88–93. PubMed Abstract | Publisher Full Text\n\nMcCully ML, Ladell K, Hakobyan S, et al.: Epidermis instructs skin homing receptor expression in human T cells. Blood. 2012; 120(23): 4591–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCalzascia T, Masson F, Di Berardino-Besson W, et al.: Homing phenotypes of tumor-specific CD8 T cells are predetermined at the tumor site by crosspresenting APCs. Immunity. 2005; 22(2): 175–84. PubMed Abstract | Publisher Full Text\n\nMikhak Z, Strassner JP, Luster AD: Lung dendritic cells imprint T cell lung homing and promote lung immunity through the chemokine receptor CCR4. J Exp Med. 2013; 210(9): 1855–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerad M, Sathe P, Helft J, et al.: The dendritic cell lineage: ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting. Annu Rev Immunol. 2013; 31: 563–604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMolenaar R, Greuter M, van der Marel AP, et al.: Lymph node stromal cells support dendritic cell-induced gut-homing of T cells. J Immunol. 2009; 183(10): 6395–402. PubMed Abstract | Publisher Full Text\n\nCording S, Wahl B, Kulkarni D, et al.: The intestinal micro-environment imprints stromal cells to promote efficient Treg induction in gut-draining lymph nodes. Mucosal Immunol. 2014; 7(2): 359–68. PubMed Abstract | Publisher Full Text\n\nPradeu T, Jaeger S, Vivier E: The speed of change: towards a discontinuity theory of immunity? Nat Rev Immunol. 2013; 13(10): 764–9. PubMed Abstract | Publisher Full Text\n\nFletcher AL, Acton SE, Knoblich K: Lymph node fibroblastic reticular cells in health and disease. Nat Rev Immunol. 2015; 15(6): 350–61. PubMed Abstract | Publisher Full Text\n\nQi H, Kastenmüller W, Germain RN: Spatiotemporal basis of innate and adaptive immunity in secondary lymphoid tissue. Annu Rev Cell Dev Biol. 2014; 30: 141–67. PubMed Abstract | Publisher Full Text\n\nFasnacht N, Huang HY, Koch U, et al.: Specific fibroblastic niches in secondary lymphoid organs orchestrate distinct Notch-regulated immune responses. J Exp Med. 2014; 211(11): 2265–79. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOnder L, Narang P, Scandella E, et al.: IL-7-producing stromal cells are critical for lymph node remodeling. Blood. 2012; 120(24): 4675–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalhotra D, Fletcher AL, Turley SJ: Stromal and hematopoietic cells in secondary lymphoid organs: partners in immunity. Immunol Rev. 2013; 251(1): 160–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIngulli E, Mondino A, Khoruts A, et al.: In vivo detection of dendritic cell antigen presentation to CD4+ T cells. J Exp Med. [Accessed December 30, 2015] 1997; 185(12): 2133–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHickman HD, Takeda K, Skon CN, et al.: Direct priming of antiviral CD8+ T cells in the peripheral interfollicular region of lymph nodes. Nat Immunol. 2008; 9(2): 155–65. PubMed Abstract | Publisher Full Text\n\nBarral P, Polzella P, Bruckbauer A, et al.: CD169+ macrophages present lipid antigens to mediate early activation of iNKT cells in lymph nodes. Nat Immunol. 2010; 11(4): 303–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGonzalez SF, Lukacs-Kornek V, Kuligowski MP, et al.: Capture of influenza by medullary dendritic cells via SIGN-R1 is essential for humoral immunity in draining lymph nodes. Nat Immunol. 2010; 11(5): 427–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWoodruff MC, Heesters BA, Herndon CN, et al.: Trans-nodal migration of resident dendritic cells into medullary interfollicular regions initiates immunity to influenza vaccine. J Exp Med. 2014; 211(8): 1611–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBecattini S, Latorre D, Mele F, et al.: T cell immunity. Functional heterogeneity of human memory CD4+ T cell clones primed by pathogens or vaccines. Science. 2015; 347(6220): 400–6. PubMed Abstract | Publisher Full Text\n\nKuraoka M, Schmidt AG, Nojima T, et al.: Complex Antigens Drive Permissive Clonal Selection in Germinal Centers. Immunity. 2016; 44(3): 542–552. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGagliani N, Amezcua Vesely MC, Iseppon A, et al.: Th17 cells transdifferentiate into regulatory T cells during resolution of inflammation. Nature. 2015; 523(7559): 221–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaldi A, Carneiro J, Chaouiya C, et al.: Diversity and plasticity of Th cell types predicted from regulatory network modelling. PLoS Comput Biol. [Accessed January 6, 2016] 2010; 6(9): e1000912. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaumjohann D, Ansel KM: MicroRNA-mediated regulation of T helper cell differentiation and plasticity. Nat Rev Immunol. [Accessed January 6, 2016] 2013; 13(9): 666–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang C, Collins M, Kuchroo VK: Effector T cell differentiation: are master regulators of effector T cells still the masters? Curr Opin Immunol. 2015; 37: 6–10. PubMed Abstract | Publisher Full Text\n\nCannons JL, Lu KT, Schwartzberg PL: T follicular helper cell diversity and plasticity. Trends Immunol. 2013; 34(5): 200–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaz A, Groves P, Buttigieg K, et al.: Quantitative assessment of the functional plasticity of memory CD8+ T cells. Eur J Immunol. 2016; 46(4): 863–73. PubMed Abstract | Publisher Full Text\n\nSeifert M, Przekopowitz M, Taudien S, et al.: Functional capacities of human IgM memory B cells in early inflammatory responses and secondary germinal center reactions. Proc Natl Acad Sci U S A. 2015; 112(6): E546–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCerwenka A, Lanier LL: Natural killer cell memory in infection, inflammation and cancer. Nat Rev Immunol. 2016; 16(2): 112–23. PubMed Abstract | Publisher Full Text\n\nGilbert CD, Li W: Top-down influences on visual processing. Nat Rev Neurosci. 2013; 14(5): 350–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBensmaia SJ, Miller LE: Restoring sensorimotor function through intracortical interfaces: progress and looming challenges. Nat Rev Neurosci. 2014; 15(5): 313–25. PubMed Abstract | Publisher Full Text\n\nMatzinger P: The danger model: a renewed sense of self. Science. 2002; 296(5566): 301–5. PubMed Abstract | Publisher Full Text\n\nMatzinger P: Tolerance, danger, and the extended family. Annu Rev Immunol. 1994; 12: 991–1045. PubMed Abstract | Publisher Full Text\n\nSchwartz M, Kipnis J: Protective autoimmunity: regulation and prospects for vaccination after brain and spinal cord injuries. Trends Mol Med. [Accessed December 14, 2015] 2001; 7(6): 252–8. PubMed Abstract | Publisher Full Text\n\nSchwartz M, Shaked I, Fisher J, et al.: Protective autoimmunity against the enemy within: fighting glutamate toxicity. Trends Neurosci. 2003; 26(6): 297–302. PubMed Abstract | Publisher Full Text\n\nSchwartz M, Baruch K: Breaking peripheral immune tolerance to CNS antigens in neurodegenerative diseases: Boosting autoimmunity to fight-off chronic neuroinflammation. J Autoimmun. 2014; 54: 8–14. PubMed Abstract | Publisher Full Text\n\nZiv Y, Ron N, Butovsky O, et al.: Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood. Nat Neurosci. 2006; 9(2): 268–75. PubMed Abstract | Publisher Full Text\n\nLavin Y, Winter D, Blecher-Gonen R, et al.: Tissue-resident macrophage enhancer landscapes are shaped by the local microenvironment. Cell. 2014; 159(6): 1312–1326. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMortha A, Chudnovskiy A, Hashimoto D, et al.: Microbiota-dependent crosstalk between macrophages and ILC3 promotes intestinal homeostasis. Science. 2014; 343(6178): 1249288. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSica A, Mantovani A: Macrophage plasticity and polarization: in vivo veritas. J Clin Invest. 2012; 122(3): 787–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwartz M, Kipnis J, Rivest S, et al.: How do immune cells support and shape the brain in health, disease, and aging? J Neurosci. 2013; 33(45): 17587–96. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "16829",
"date": "06 Oct 2016",
"name": "Irun R. Cohen",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper by Serge Nataf, Clonal selection versus clonal cooperation: the integrated perception of immune objects, is aptly named: here Nataf draws an analogy between the immune system (IS) and the central nervous system (CNS); both systems, he points out, are cognitive in the sense that they construct their objects of attention through the integration of discrete features perceived individually. The CNS constructs specific visual objects by integrating information between networks of neurons that gather discrete and separate information about shape, location, color and motion; similarly, the IS constructs immune objects by integrating discrete information about antigens, cytokines, chemokines, and other signals obtained separately by various cell types such as macrophages, dendritic cells, and B and T lymphocytes. Integration of information between different cells and cell types requires cooperation between the interacting cells; thus the IS, like the CNS, makes functional decisions based on clonal cooperation and not on clonal competition for survival of the fittest. The cognitive analogy is presented in a clear and stimulating manner. However, I think that Serge Nataf could increase the impact of his CNS-IS discussion by sharpening or expanding some additional points:\nFunctional development requires somatic experience: Both the CNS and the IS combine individual somatic experience with innate programming acquired through evolution of the species. In fact, the CNS and the IS are the only two mammalian systems that require post-natal, individual contact with the world to self-organize their mature structures and functions (see Cohen I.R, (2000)); an IS deprived of immune experience remains undeveloped as does a CNS deprived of formative stimulation.\n\nHomunculi direct attention to environmental information that serves adaptive fitness: Nataf mentions briefly that both the CNS and the IS use homunculi – internal representations of selected features of the self and the outside world; he might add that such homunculi exist to maximize evolutionary fitness. The human brain, for example, is born hard-wired to be attracted to human faces; this primitive homuncular sense of facial recognition is sharpened by individual post-natal experience with real persons, beginning with mother and leading to success in interpersonal bonding, friendship and other important relationships, and to the capacity to communicate by facial expressions.\n\nThe immunological homunculus, which is encoded in innate immune receptors and in shared autoantibody repertoires (Madi A, et al. (2015)) and in public T cell receptor repertoires (Madi A, et al. (2014)) is much less appreciated. But, like the Neurological Homunculus, these congenital and acquired receptors are likely to enhance fitness, for example, by enabling the IS to better preserve and heal the self and its microbiome while protecting the body from pathogenic invaders from without and from tumor cells from within.\n\nImportant points of difference between the CNS and IS would include hard wiring compared to cell migration, reaction times of milliseconds to hours and days, markedly different cell turnover and differentiation rates, and others; both systems create and integrate their perceived objects in markedly different ways. But, as Nataf points out so convincingly, integrated cooperation is the hallmark of both systems.",
"responses": []
},
{
"id": "16681",
"date": "12 Oct 2016",
"name": "Alfred I. Tauber",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSerge Nataf has admirably schematized the “cognitive paradigm,” so aptly coined by Irun Cohen 25(!) years ago (see his citations in the accompanying review). The immune system is, indeed, a “mobile” brain (Fridman 1991) and the parallels between the immune and nervous systems is far from metaphorical or idealized modeling: their inter-connections include the origins of their respective developments, that continue with intimate anatomic relationships, as well as shared messengers and receptors. Indeed, psychoneuroimmunology comprises a large research field, whose wealth of studies have confirmed the close evolutionary histories of what are, in essence, similar information processing systems (Ader, Felten and Cohen 1991; Ader 2006; Orosz 2001; Forrest and Hofmeyr 2001). (Note, instead of using the visual system as his example, Nataf might have shown the obvious similarities in the molecular-sensing perception of taste or smell to make the same point about perception and, in that example, emphasized the similarities of structural correspondence and subsequent activation employed by both the immune and taste systems.)\nTwo historical notes:\nFirst, while Nataf credits Jerne with introducing cognitive terminology to immunology (citing the 1974 idiotypic network paper), the origins of immunity as cognitive appears earlier in the cybernetics craze of the early 1950s and Burnet’s musings a decade later (Tauber 1994). (Although Burnet first used \"cognate\" in passing in 1959 [Burnet 1959, p.70], by 1962, he invoked an analogy with language to account for antibody selection [Burnet 1962, p. 94-5].) But historical primacy is not the matter of interest regarding Jerne’s seminal paper, but rather his introduction of how the context of antigen presentation determined the immune response. According to his hypothesis, the disruption of the network by antigen, and more to the point, the extent of that interruption in terms of breaking the inter-connections of the lattice-like structure, determined the immunogenicity of the introduced substance. That fundamental idea is the source of Matzinger’s Danger Theory and Pradeu’s more recent continuity/discontinuity theory.\nThe second historical note is that an important chapter of the cognitive paradigm’s history must include the work of Antonio Coutinho (1991; 1995; 2003), Francisco Varela (Varela et al., 1988; 1991, 1993; Varela and Coutinho 1991), John Stewart (1992; 1994a; 1994b; Stewart and Varela 1989), and Nelson Vaz (2016; Vaz et al., 2006). These publications developed Jerne’s cognitive point in its full theoretical array and attempted to draw both theoretical and practical parallels between the nervous and immune systems. Those studies comprise the rich mulch for the emergence of our current thinking about how immunity may be properly regarded as an information processing faculty and as such shares organizational (and ultimately regulatory) characteristics with the brain.\nIn terms of going beyond Nataf’s schema as an educational tool (which it certainly is), a deeper assessment must be made as to how well modelers have applied the integrative principles employed here. In other words, to the extent systems biology has offered models of immune responses or steady-state conditions, has the hierarchical structure employed by Nataf been utilized and if so, what insights does such an approach offer? That assessment would be useful not only in terms of confirming Nataf’s schema, but perhaps modelers might develop this approach for their own efforts. A review of this issue, as well as a critical discussion about the cognitive paradigm applied to immunology, is found elsewhere (Tauber 1997; 2013; 2017).",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2226
|
https://f1000research.com/articles/5-2170/v1
|
02 Sep 16
|
{
"type": "Study Protocol",
"title": "Open discectomy vs microdiscectomy for lumbar disc herniation - a protocol for a pragmatic comparative effectiveness study",
"authors": [
"Andreas Sørlie",
"Sasha Gulati",
"Charalampis Giannadakis",
"Sven M. Carlsen",
"Øyvind Salvesen",
"Øystein P. Nygaard",
"Tore K. Solberg",
"Charalampis Giannadakis",
"Sven M. Carlsen",
"Øyvind Salvesen",
"Øystein P. Nygaard"
],
"abstract": "Introduction: Since the introduction of lumbar microdiscectomy in the 1970’s, many studies have attempted to compare the effectiveness of this method with that of standard open discectomy with conflicting results. This observational study is designed to compare the relative effectiveness of microdiscectomy (MD) with open discectomy (OD) for treating lumbar disc herniation, -within a large cohort, recruited from daily clinical practice. Methods and analysis:\n\nThis study will include patients registered in the Norwegian Registry for Spine Surgery (NORspine). This clinical registry collects prospective data, including preoperative and postoperative outcome measures as well as individual and demographic parameters. The primary outcome is change in Oswestry disability index between baseline and 12 months after surgery. Secondary outcome measures are improvement of leg pain and changes in health related quality of life measured by the Euro-Qol-5D between baseline and 12 months after surgery, complications to surgery, duration of surgical procedures and length of hospital stay.",
"keywords": [
"Microdiscectomy",
"Discectomy",
"multicenter",
"comparative",
"Effectiveness",
"Registry",
"pragmatic (study protocol)",
"Propensity match"
],
"content": "Background\n\nLumbar disc herniation is a common cause of sciatic pain and functional disability. Although most patients are relieved from their symptoms without surgical treatment, there is consensus for operating on selected patients with persistent radicular pain after 2–6 months (Atlas et al., 927–35; Schoenfeld & Bono, 1963–70; Weinstein et al., 2789–800). Surgical discectomy gives earlier relief of symptoms, enabling patients to return to their work and other daily activities more rapidly. In cases where the pain is incapacitating and the patient is bedridden with strong pain medication or in the cases of progressive paresis, there are sometimes indications for earlier surgical intervention. The aim is to relieve the patient of the pain and to prevent late sequela, like permanent paresis and neuropathic pain. Moreover, prolonged sick leave may lead to undesirable lifestyle changes and reduce motivation to return to work (Sieurin et al., 50–56).\n\nIn 1977, Yasargil and Caspar independently introduced the technique of microdiscectomy for treating lumbar disc herniation (Yasargil, 81) (Caspar et al., 78–86). This technique offers better visual control of the operation field, through less traumatic and smaller incisions, compared to the standard open discectomy. The role of other treatment options such as minimally invasive discectomy (MID), chemonucleolysis and endoscopic discectomy is still unclear (Rasouli et al., CD010328) (Gibson & Waddell, 1735–47) and both open discectomy and microdiscectomy are still considered the best surgical treatment options (Gibson & Waddell, 1735–47) and are the most commonly used treatment modalities today.\n\nPrevious randomized trials have been small single centre studies and have been unable to demonstrate any difference between the two treatment modalities (Katayama et al., 344–47; Tullberg et al., 24–27). In one recent prospective non-randomised multicenter trial of 261 patients, they found significant better improvement of radicular pain at 12 months follow up after open discectomy in comparison to microdiscectomy. For all other outcome parameters, there were no significant differences between the groups (Porchet et al., 360–66). However, the two treatment cohorts were not matched and uneven with respect to risk factors that might influence the outcome. A systematic review done by Gibson and Waddell in 2007, found no significant difference in outcome between the two treatment modalities and they concluded that even though open discectomy remains the “standard”, further studies comparing these two surgical methods are warranted. (Gibson & Waddell, 1735–47).\n\nThe current study is a large multicenter observational study comparing the relative effectiveness of the two treatments. Data are obtained as part of daily clinical practice at several institutions, resulting in a high external validity. Moreover, the size of the study allows for propensity matching, making the two groups comparable in most aspects for a close approximation to a randomized controlled trial.\n\nThis observational study is designed to compare the relative effectiveness of discectomy with or without visual enhancement for treating lumbar disc herniation. We will use the term “open discectomy” for discectomy done without the use of visual enhancement, while the term “microdiscectomy” entails the use of visual enhancement like a microscope or loupes.\n\nThere are some variations in the surgical procedure, depending on peroperative findings as well as surgeon preferences. One such variation is whether the disc space is entered during the operation. In this study, the term discectomy includes procedures were sequestrectomy is done with or without entering the disc space. Variations of the surgical procedure (in both groups), such as discectomy vs. sequestrectomy and concomitant decompression of the nerve root due to recess- or foraminal stenosis will be reported (See Table 2 in Supplementary material).\n\n\nAims of the study\n\nThe primary aim of this study is to compare the effectiveness of discectomy with or without visual enhancement (i.e. microdiscectomy vs. open discectomy) for treating lumbar disc herniation.\n\n\nMethods and materials\n\nStudy population: Data for this cohort study have been collected through NORspine which is a Norwegian national comprehensive clinical registry for quality control and research of surgical intervention in the spine. Participation in the registration is voluntary, however patients are recommended to participate for contributing to the completeness of the registry. Patients receive the same treatment irrespective of their participation in the registry.\n\nPatients operated on between October 2006 and the end of May 2014 will be screened for study eligibility. In the registry, the follow-up time of the operation (at baseline) is 3 and 12 months.\n\nInclusion criteria:\n\n1. Included in the NORspine registry\n\n2. Surgery for herniated lumbar disc disease using open discectomy or microdiscectomy with preservation of midline structures (spinous process and ligaments).\n\nExclusion criteria:\n\n1. Operated in > 1 level.\n\n2. Previous operations in the lumbar spine.\n\n3. Patients with deformities in the lumbar spine (spondylolisthesis or scoliosis).\n\n4. If intervention included more comprehensive surgery like laminectomy or fusion.\n\n5. Other minimally invasive procedures.\n\n6. Far lateral approaches for lateral disc herniations.\n\nOn admission for surgery, patients complete the baseline questionnaire, which includes questions about demographic and lifestyle issues in addition to the outcome measures and duration of symptoms. Information about marital status, educational level, employment status, body mass index, and tobacco-smoking is available in the NORspine registry. The surgeon records data concerning diagnosis, comorbidity, American Society of Anesthesiologists (ASA) grade, image findings, treatment, use of prophylactic antibiotics and peroperative complications. Duration of the surgical procedure and hospital stay are recorded by hospital staff (trained nurse or health secretary). A questionnaire is distributed by regular mail 3 and 12 months after surgery, completed at home by the patient and returned to the central registry unit of the NORspine, without involving the treating hospitals. According to a standardized set of questions patients report postoperative complications having occurred within 3 months of follow-up. Non respondents receive one reminder with a new copy of the questionnaire. The response rates to the registry after 1 year for the relevant group of patients is around 70% (Figure 1). Figure 1 shows the patient population before the statistical analysis will commence.\n\nN refers to surgical procedures.\n\nEthics and dissemination: NORspine has been evaluated and approved by the Norwegian Data Protection Authority. All participants have provided written informed consent that the data collected in NORspine can be used for research purposes and the results will be disseminated through peer-reviewed publications. The study has been approved by the regional committee for medical and health research in central Norway (REK central) 2016/840.\n\nPrimary outcome measures: The primary outcome measure is change in functional outcome defined by Oswestry disability index (ODI) between baseline and follow-up at 12 months (mean change of ODI and proportion of patients achieving the minimally clinical important difference (MCID) between the two groups).\n\nThe functional outcome is measured with V.2.0 of ODI and translated into Norwegian and tested for psychometric properties as outlined by Grotle et al. (Grotle et al., 241–47). ODI is one of the principal condition-specific outcome measures used in the management of spinal disorders. ODI contains 10 questions on limitation of activities of daily living. Each variable is rated on a 0 – 5 point scale, summarised and converted into a percentage score, ranging from 0 to 100 (0= no disability).\n\nThere are great variations in the estimated values for MCID after spine surgery and the value ranges from 8 – 15 in the literature (Hellum et al., d2786; Nerland et al., h1603; Ostelo et al., 90–94). MCID is by the definition the minimal value of improvement in a measurement that exceeds the normal statistical variation and also is experienced by the patient as a definite improvement. In most cases the value of MCID does not however represent the desired effect of the treatment, i.e. it is not a criterium for success. Solberg et al. established anchor based success criteria (Solberg et al., 196–201) after discectomy as change in ODI ≥ 20 and change in NRS leg pain ≥3,5. Both MCID and cutoff values for success are useful tools for evaluating patient outcome and can be used to compare proportions of patients responding to the treatment. To estimate the proportion of responders to the treatment, we define an improvement of ODI ≥ 10 and NRS leg pain ≥ 2 as cutoff of minimally important clinical differences (Ostelo et al., 90–94) (Nerland et al., h1603) (Brox et al., 145–55) (Hellum et al., d2786) .\n\nSecondary outcome measures: Secondary measures are:\n\n1. Mean changes in health related quality of life measured with the EQ-5D between baseline and 12 months follow-up.\n\n2. Mean improvement of NRS leg pain between baseline and follow-up at 12 months.\n\n3. Mean improvement in back pain using the Numeric Rating Scale (NRS).\n\n4. Duration of the surgical procedure.\n\n5. Duration of hospital stay.\n\n6. Perioperative complications.\n\n7. Postoperative complications.\n\nEQ-5D is a generic and preference-weighted measure of HRQL. The Norwegian version of EQ-5D has shown good psychometric properties and has been validated for patient populations similar to that in our study (Solberg et al., 1000–07). EQ-5D evaluates five dimensions: mobility, self-care, activities of daily living, pain, and anxiety and/or depression. For each dimension, the patient describes three possible levels of problems (non, mild-to-moderate and severe). This descriptive system therefore contains 243 (35) combinations or index values for health status. EQ-5D total score ranges from -0.6 to 1, where 1 corresponds to perfect health and negative values are considered worse than death (Dolan et al., 141–54).\n\nLeg pain will be assessed by the Numeric Rating Scale (NRS), ranging from 0–10, where 0 = no pain and 10 = worst pain.\n\nPerioperative complications are reported at the time of inclusion (immediately after surgery) and include dural tear, nerve root injury, bleeding requiring transfusion, respiratory or cardiovascular complications and anaphylaxis.\n\nPostoperative complications are registered by the patient on the follow up questionnaire after 3 months and includes wound infection, deep venous thrombosis, pulmonary embolism, pneumonia and urinary tract infection. We also define reoperation within 3 months as a complication.\n\n\nSurgical procedures\n\nSince this is a multicenter trial, variations in the surgical management and the surgical procedures can only be described in general terms and in accordance with the data collected in the NORspine registry. The microsurgical discectomy is well described and involves preoperative fluoroscopy for detection of the target level, paramedian or median incision of about 3–6 cm, straight or curved opening of the paravertebral muscular fascia, subperiosteal release of the paravertebral musculature from the spinous process and basal lamina above and below the target disc-level. Caspar self-retaining retractors and a microscope or loupes are introduced. In most cases a flavectomy and arcotomy of the lamina above the disc-level is done. Careful mobilization of the dural sac and the nerve-root medially, before evacuating the herniated disc. This might involve entering the disc space, or just removing a free sequestrated disc fragment (sequestrectomy). The traditional open discectomy did not involve retractors which minimizes the incision to unilateral muscular dissection. However, many institutions that perform standard open discectomy also use equivalents to the caspar retractors. In which case the procedure is in principle the same as described for microdiscectomy, except regarding the use of microscope or other visually enhancing tools (like loupes) and may require a larger incision and more soft tissue damage.\n\n\nStatistical analyses\n\nThis study will test the equivalence of the clinical effectiveness of the two surgical techniques. Case analysis will be done using mixed linear model analysis in both the aggregate cohort and a propensity matched cohort. The minimal clinical important difference for change in the mean ODI score is considered to be in the range of 10 points. If mean changes of ODI is < 10, the treatments are considered equal with respect to effectiveness. Since there is no clear consensus on how large the MCID should be between two treatment groups and as the MCID also describes effects of interventions on an individual level, we would like to present the proportion of patients having MCID as result of the treatment – in both the aggregate cohort and the propensity matched cohort. In the analysis of primary and secondary outcome measures, adjustments for age, body mass index, and preoperative ODI, as well as smoking habits will be done. Statistical significance level is defined as p <0.05 with no adjustments made for multiple comparisons. Baseline and follow-up measurements will be assumed to be normally distributed provided this assumption is confirmed by Q-Q plots. To evaluate the magnitude of change in EQ-5D score, we will estimate effect sizes according to the method of Kazis. (Kazis et al., S178–S189). An effect size of 0.8 or more is considered to be large. In the mixed model patients will not be excluded from the analysis, if the variable is missing at some (but not all) time points after baseline. This strategy is in line with a study showing that it is not necessary to handle missing data using multiple imputations before performing a mixed model analysis on longitudinal data.(Twisk et al., 1022–28). In the additional analyses (categorical data at three months’ follow-up), we will not replace missing data. Continuous variables will be analysed using an unpaired two tailed t test for normally distributed data, and Mann- Whitney U test for skewed distribution. A X2 analysis will be used to compare discrete variables. The content of tables and figures (see Supplementary material) are predefined before the statistical analyses are done, and no information will be deleted when results are known. We do not plan any additional exploratory statistical analyses.\n\nTo achieve the closest approximate to a randomized clinical trial, we will use matching approach technique of using propensity scores, as opposed to stratification or regression adjustment. It provides the greatest balance between the two treatment groups (Hemmila et al., 939–45) (Austin et al., 734–53). We will generate propensity scores for surgical technique using logistic regression and adjusting for baseline covariates that could influence treatment outcomes, including age, sex, life partner, comorbidity, body mass index, smoking, educational level, and preoperative ODI score. All covariates are entered into a logistic regression analysis, and we will fit a maximum likelihood model based on these covariates as predictors of surgical technique. The coefficients for these predictors of surgical technique are used to calculate a propensity score of 0 to 1 for each patient. Based on the calculated propensity scores, two evenly matched groups will be formed for surgical technique using a matching algorithm with the common caliper set at 0.010. This dataset will be referred to as the “propensity matched cohort.” We will analyze continuous variables using a related samples two tailed t test for data with a normal distribution and continuous data exhibiting a skewed distribution using the Wilcoxon matched pair signed rank test. We will use the McNemar’s test to compare discrete variables.\n\n\nStudy limitations\n\nThe main limitation of this study, as its purpose is to compare two different treatment modalities, is that it is not a randomized trial. Rather than comparing the efficacy of the treatment, we will focus on the effectiveness, and the findings might entail other treatment related differences than that of the surgical procedure alone. However, the use of propensity matched groups will minimize these potential differences.\n\nFor the standard discectomies, we cannot differentiate between operations done with unilateral or bilateral muscular dissection.\n\nThe inclusion rate to the NORspine registry for lumbar herniated disc procedures is around 65%. The inclusion rate is closely monitored through comparing registered patients with data from the Norwegian Patient Registry (NPR) (where all patients operations are registered for performance-based financing). Loss to follow up may be approximately 30%. However, a previous study has shown that nonresponders have the same outcome after surgery as those who respond to the follow up questionnaires (Solberg et al., 56–63). These factors might nevertheless limit the validity of our findings.\n\n\nConclusion\n\nThis is a protocol for an observational study designed to compare the relative effectiveness of microdiscectomy with open discectomy for treating lumbar disc herniation. The study is based on data from the NORspine registry collected from 30 different institutions in Norway from October 2006 – 12. May 2014. We have discussed the details of the clinical registry and patient enrolment as well as the planned statistical analysis.",
"appendix": "Author contributions\n\n\n\nAndreas Sørlie: main author\n\nSasha Gulati: co-author, main coordinator regarding statistical analysis\n\nCharalampis Giannadakis: co-author - contributor to statistical analysis\n\nSven M. Carlsen: co-author and contributor to statistical analysis\n\nØyvind Salvesen: co-author\n\nØystein P Nygaard: - co author\n\nTore Solberg: co-author and contributor to statistical analysis\n\nAll authors agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nTable 1: Baseline characteristics for both treatment groups.\n\nClick here to access the data.\n\nTable 2: Procedural differences and differences in peroperative- and postoperative complications between the two groups.\n\nClick here to access the data.\n\nTable 3: Changes in ODI and EQ-5D between baseline and one year after the operation within each treatment group for both the aggregate cohort and the matched cohort.\n\nClick here to access the data.\n\nTable 4: Complete case analyses of categorical outcome measures in the aggregate cohort and propensity matched cohort. MCID is defined as dODI ≥ 10 and dNRS ≥ 2.\n\nClick here to access the data.\n\n\nReferences\n\nAtlas SJ, Keller RB, Wu YA, et al.: Long-term outcomes of surgical and nonsurgical management of sciatica secondary to a lumbar disc herniation: 10 year results from the maine lumbar spine study. Spine (Phila Pa 1976). 2005; 30(8): 927–35. PubMed Abstract | Publisher Full Text\n\nAustin PC, Grootendorst P, Anderson GM: A comparison of the ability of different propensity score models to balance measured variables between treated and untreated subjects: a Monte Carlo study. Stat Med. 2007; 26(4): 734–53. PubMed Abstract | Publisher Full Text\n\nBrox JI, Reikerås O, Nygaard Ø, et al.: Lumbar instrumented fusion compared with cognitive intervention and exercises in patients with chronic back pain after previous surgery for disc herniation: a prospective randomized controlled study. Pain. 2006; 122(1–2): 145–55. PubMed Abstract | Publisher Full Text\n\nCaspar W, Campbell B, Barbier DD, et al.: The Caspar microsurgical discectomy and comparison with a conventional standard lumbar disc procedure. Neurosurgery. 1991; 28(1): 78–86, discussion 86–7. PubMed Abstract | Publisher Full Text\n\nDolan P, Gudex C, Kind P, et al.: The time trade-off method: results from a general population study. Health Econ. 1996; 5(2): 141–54. PubMed Abstract | Publisher Full Text\n\nGibson JN, Waddell G: Surgical interventions for lumbar disc prolapse: updated Cochrane Review. Spine (Phila Pa 1976). 2007; 32(16): 1735–47. PubMed Abstract | Publisher Full Text\n\nGrotle M, Brox JI, Vøllestad NK: Cross-cultural adaptation of the Norwegian versions of the Roland-Morris Disability Questionnaire and the Oswestry Disability Index. J Rehabil Med. 2003; 35(5): 241–47. PubMed Abstract | Publisher Full Text\n\nHellum C, Johnsen LG, Storheim K, et al.: Surgery with disc prosthesis versus rehabilitation in patients with low back pain and degenerative disc: two year follow-up of randomised study. BMJ. 2011; 342: d2786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHemmila MR, Birkmeyer NJ, Arbabi S, et al.: Introduction to propensity scores: A case study on the comparative effectiveness of laparoscopic vs open appendectomy. Arch Surg. 2010; 145(10): 939–45. PubMed Abstract | Publisher Full Text\n\nKatayama Y, Matsuyama Y, Yoshihara H, et al.: Comparison of surgical outcomes between macro discectomy and micro discectomy for lumbar disc herniation: a prospective randomized study with surgery performed by the same spine surgeon. J Spinal Disord Tech. 2006; 19(5): 344–47. PubMed Abstract | Publisher Full Text\n\nKazis LE, Anderson JJ, Meenan RF: Effect sizes for interpreting changes in health status. Med Care. 1989; 27(3 Suppl): S178–S189. PubMed Abstract | Publisher Full Text\n\nNerland US, Jakola AS, Solheim O, et al.: Minimally invasive decompression versus open laminectomy for central stenosis of the lumbar spine: pragmatic comparative effectiveness study. BMJ. 2015; 350: h1603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOstelo RW, Deyo RA, Stratford P, et al.: Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976). 2008; 33(1): 90–94. PubMed Abstract | Publisher Full Text\n\nPorchet F, Bartanusz V, Kleinstueck FS, et al.: Microdiscectomy compared with standard discectomy: an old problem revisited with new outcome measures within the framework of a spine surgical registry. Eur Spine J. 2009; 18(Suppl 3): 360–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRasouli MR, Rahimi-Movaghar V, Shokraneh F, et al.: Minimally invasive discectomy versus microdiscectomy/open discectomy for symptomatic lumbar disc herniation. Cochrane Database Syst Rev. 2014; (9): CD010328. PubMed Abstract | Publisher Full Text\n\nSchoenfeld AJ, Bono CM: Does surgical timing influence functional recovery after lumbar discectomy? A systematic review. Clin Orthop Relat Res. 2015; 473(6): 1963–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSieurin L, Josephson M, Vingård E: Positive and negative consequences of sick leave for the individual, with special focus on part-time sick leave. Scand J Public Health. 2009; 37(1): 50–56. PubMed Abstract | Publisher Full Text\n\nSolberg T, Johnsen LG, Nygaard ØP, et al.: Can we define success criteria for lumbar disc surgery? : estimates for a substantial amount of improvement in core outcome measures. Acta Orthop. 2013; 84(2): 196–201. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSolberg TK, Olsen JA, Ingebrigtsen T, et al.: Health-related quality of life assessment by the EuroQol-5D can provide cost-utility data in the field of low-back surgery. Eur Spine J. 2005; 14(10): 1000–07. PubMed Abstract | Publisher Full Text\n\nSolberg TK, Sørlie A, Sjaavik K, et al.: Would loss to follow-up bias the outcome evaluation of patients operated for degenerative disorders of the lumbar spine? Acta Orthop. 2011; 82(1): 56–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTullberg T, Isacson J, Weidenhielm L: Does microscopic removal of lumbar disc herniation lead to better results than the standard procedure? Results of a one-year randomized study. Spine (Phila Pa 1976). 1993; 18(1): 24–27. PubMed Abstract | Publisher Full Text\n\nTwisk J, de Boer M, de Vente W, et al.: Multiple imputation of missing values was not necessary before performing a longitudinal mixed-model analysis. J Clin Epidemiol. 2013; 66(9): 1022–28. PubMed Abstract | Publisher Full Text\n\nWeinstein JN, Lurie JD, Tosteson TD, et al.: Surgical versus nonoperative treatment for lumbar disc herniation: four-year results for the Spine Patient Outcomes Research Trial (SPORT). Spine (Phila Pa 1976). 2008; 33(25): 2789–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYasargil MG: Microsurgical Operation of Herniated Lumbar disc. Adv Neurosurgery. 1971; 4: 81. Publisher Full Text"
}
|
[
{
"id": "16070",
"date": "21 Sep 2016",
"name": "Terje Sundstrøm",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have designed a study to investigate the difference in outcome between microdiscectomy and open discectomy as treatments for lumbar disc herniation. Microdiscectomy refers to a discectomy with the use of magnification (microscope or loupes) and open discectomy refers to a discectomy without the use of magnification. This is a real-world, large, observational study using data from 2006 to 2014 from the national spine registry in Norway. The study is well designed, will have a high degree of external validity and can approximate the strength of a randomized controlled trial; the authors will include propensity score matching to reduce the potential bias of a range of baseline co-variables.\n\nPrevious retrospective and prospective observational studies of microdiscectomy versus open discectomy have landed on both sides of the fence. However, three randomized controlled trials have not demonstrated any superiority between the two operative techniques 1-3.\n\nIn an updated Cochrane review from 2007 4, it was argued that microdiscectomy and open discectomy are both gold standard treatments for symptomatic lumbar disc herniation. Nonetheless, contemporary use of microdiscectomy or open discectomy is probably relatively independent of the accumulated scientific evidence, and to a larger extent reliant on the training and expertise of the surgeon, as well as the availability and quality of visual augmentation systems. We feel that this study is a little like documenting the need for binoculars in birdwatching. An experienced birdwatcher may identify nearly every species without binoculars, but for some reason they all use binoculars; they can simply see much better. We can agree that there is still a need for more and better evidence on the relative clinical outcomes and cost-effectiveness of microdiscectomy versus open discectomy, but microscopes significantly augment vision and – let us not forget – greatly facilitate teaching. For these latter reasons, most (if not all) major spine units worldwide utilize some kind of magnification.\n\nIt would be interesting to know if the relative number of microdiscectomies and open discectomies varied over time, or if its simply a gradual conversion from open surgery to microsurgery. Whatever the outcome of this study may be, we are concerned about the interpretation, because we do not know why some surgeons chose not to use any kind of visual magnification. In the study period in Norway, most of the open discectomies were probably done by experienced surgeons (and some by reckless individuals), whereas many of the microdiscectomies were probably done by less experienced surgeons or surgeons in training. Including the identity of the surgeons in the data analysis could help unmask possible differences more attributable to the surgeon than the surgical technique per se.\n\nVisual augmentation devices are widely used, and microscopes are unsurpassable in the training of young surgeons. From this, and the accumulated literature in this field showing similar or minor differences, it is in our opinion unlikely that any positive or negative result of microdiscectomy will affect the way surgeons choose to operate lumbar disc herniations.",
"responses": [
{
"c_id": "2203",
"date": "13 Oct 2016",
"name": "Andreas Sørlie",
"role": "Author Response",
"response": "Thank you for the referee report. We appreciate your concern regarding the interpretation of the results. One challenge in this regard is that the different centers have different approaches according to surgeon preference, and that the technique used might be the best (whichever technique is used) in the hands of that particular surgeon. However, we believe the results might influence both the chosen approach by the particular surgeon as well as the preferred handling of the group of patients by a department, as well as for training of future surgeons. Some departments in Norway perform both microdiscectomy as well as open discectomy. There seems to be a slight shift towards microdiscectomy over time, but several open discectomies are still performed every year in Norway. We can include this data in the discussion when the results are ready. Andreas Sørlie"
}
]
},
{
"id": "17020",
"date": "17 Oct 2016",
"name": "Edilson Forlin",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI believe the protocol presented is interesting and will be able to provide important information on which technique may have advantages to address discal herniation with surgical indication. The objectives and assessment criteria are clear and consistent. The assessment covers the main points of clinical practice when we think of the results.\nThe limitations and difficulties of the study are not small. The lack of control over information and aspects of surgery as pointed out by the authors can actually hinder the interpretation of results. It would be interesting to have a profile on surgeons in both groups. It seems to me microdiscectomy is preferred by younger surgeons. This and other factors related to surgery may influence the results.\nOther comparative reviews show few differences between the techniques. Several factors related to patients, indications and technique are involved and it is difficult to understand the importance of each. Furthermore preference, training and experience of each surgeon still overlaps the data from the studies. Even with great care to obtain and evaluate data such studies may have little influence on surgeons' treatment decisions.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2170
|
https://f1000research.com/articles/5-2157/v1
|
02 Sep 16
|
{
"type": "Opinion Article",
"title": "Opportunities and considerations for visualising neuroimaging data on very large displays",
"authors": [
"Matthew B. Wall",
"David Birch",
"May Y. Yong",
"David Birch",
"May Y. Yong"
],
"abstract": "Neuroimaging experiments can generate impressive volumes of data and many images of the results. This is particularly true of multi-modal imaging studies that use more than one imaging technique, or when imaging is combined with other assessments. A challenge for these studies is appropriate visualisation of results in order to drive insights and guide accurate interpretations. Next-generation visualisation technology therefore has much to offer the neuroimaging community. One example is the Imperial College London Data Observatory; a high-resolution (132 megapixel) arrangement of 64 monitors, arranged in a 313 degree arc, with a 6 metre diameter, powered by 32 rendering nodes. This system has the potential for high-resolution, large-scale display of disparate data types in a space designed to promote collaborative discussion by multiple researchers and/or clinicians. Opportunities for the use of the Data Observatory are discussed, with particular reference to applications in Multiple Sclerosis (MS) research and clinical practice. Technical issues and current work designed to optimise the use of the Data Observatory for neuroimaging are also discussed, as well as possible future research that could be enabled by the use of the system in combination with eye-tracking technology.",
"keywords": [
"Neuroimaging",
"fMRI",
"PET",
"visualisation",
"data observatory",
"display technology",
"eye-tracking",
"multiple sclerosis."
],
"content": "Introduction\n\nA natural trend in many scientific disciplines is towards greater size and complexity of the empirical data sets that are collected. This may be driven by the development of entirely new research methodologies or diagnostic tests, further refinements of existing technology (e.g. greater resolution of imaging, or higher speed of sampling), or by the incorporation of multiple measurement methods to examine a single question. In the era of ‘Big Data’ (Katal et al., 2015; Marx, 2013) some scientists are now developing specialist techniques to handle truly enormous data sets.\n\nThis trend is certainly evident in the field of neuroimaging. Functional Magnetic Resonance Imaging (fMRI) is now the workhorse method in cognitive neuroscience and can generate impressively large and complex data sets. Recent advances in fMRI acquisition software to achieve increased spatial and temporal resolution (e.g. Moeller et al., 2010) have driven a further increase in data volumes. Large scale endeavours such as the Human Connectome Project (HCP; Van Essen et al., 2013) aim to gather a variety of different data from large cohorts. The HCP is currently acquiring data using four different MRI procedures (structural, resting-state fMRI, task fMRI, and diffusion imaging), from 1200 subjects, with a sub-set also completing magnetoencephalography (MEG) and electroencephalography (EEG) scans, and a further sub-set also completing additional scans on a high-field strength (7 Tesla) MRI scanner. With additional demographic, behavioural, and questionnaire measures, the final HCP data set will be a tremendous resource, but its sheer size will require specialist methods of data-handling and analysis.\n\nThe HCP illustrates two common features of modern neuroimaging research. First is the collection of multiple types of data from a set of subjects using a single imaging modality, most commonly MRI. These may include task fMRI, resting-state fMRI, diffusion MRI, Arterial Spin Labelling (ASL), Magnetic Resonance Spectroscopy (MRS), or a number of other specialist techniques. The second is the advent of true multi-modal neuroimaging research, where combinations of two (or more) methods are used, either simultaneously or independently. Combined fMRI-EEG studies (Huster et al., 2012) combine the high spatial resolution of MRI with the high temporal resolution of EEG, often with simultaneous acquisition. MRI and MEG have also been used successfully (e.g. Carhart-Harris et al., 2016) and provide similarly complementary data, though not simultaneously. MRI and Positron Emission Tomography (PET) data can be collected independently (e.g. Colasanti et al., 2016; Rabiner et al., 2011) or simultaneously (using the new generation of combined PET/MR scanners; Bailey et al., 2015) and combine PET-derived information on neurochemistry with structural or functional MRI measures. Multimodal imaging has also begun to filter through into clinical practice with some diagnostic criteria now incorporating imaging markers of neurodegenerative disorders (e.g. in Multiple Sclerosis; Polman et al., 2011).\n\nThese multi-paradigm and multi-modality studies are of great value in providing complementary and converging evidence to characterise healthy brain function, examine various disease states, and in drug development (Matthews et al., 2011). The challenges involved in analysing and manipulating large multi-modal datasets have been partly addressed by advances in hardware and software. For example, the issue of fusing images from different modalities has largely been solved by modern software (e.g. Gunn et al., 2016) using automated co-registration algorithms that generally produce good-quality results. One remaining challenge is the provision of appropriate visualisation technologies that can provide an overview of a set of (sometimes disparate) results images, and can enable accurate interpretations to be made. Many specialised software tools now exist for visualising neuroimaging data (a comprehensive list, and a useful guide to visualisation can be found in Madan, 2015) however, their utility is necessarily constrained by the users display hardware; typically a single, or several standard desktop computer monitors. Advances in display technology have only been incompletely addressed, with most tools not optimised for larger displays, and also not incorporating modern user-interface features such as touch input. This occurs for two reasons, firstly that physical display hardware has only recently begun to support the higher resolutions required to display the higher fidelity data which are now routinely captured. Secondly the software used to display scientific data has not benefited from the revolution in distributed, cloud computing which data processing systems such as Map-Reduce and Hadoop provide (Patel et al., 2012).\n\nTo address these challenges and enable high-resolution collaborative exploration of detailed scientific data a new generation of advanced visualisation suites are being developed (Febretti et al., 2013). One example is the KPMG Data Observatory (DO) at Imperial College London. This is a panoramic display covering a 313 degree arc with a 6 m diameter, providing an immersive and collaborative space for exploration of data (see Figure 1 and Figure 2). The key differentiator of the space is its high resolution which totals 132 megapixels, in contrast with the low-resolution projector based approach of traditional CAVE systems. The system is driven by 32 rendering nodes that enable distributed analysis and rendering of data, and the display area can be flexibly configured into either a single display surface, or a number of sections displaying different information sources or applications. The key goal of the observatory is to provide a collaborative space for research teams to explore and discuss data in a visual format.\n\nProfessor Oliver Howes presenting multimodal imaging data at the MRC Clinical Sciences Centre's \"Hearts and Minds\" public engagement event, 23 June 2016. Photo Credit: Susan Watts, MRC Clinical Sciences Centre, Imperial College London, reproduced with permission.\n\nA panoramic image of the Data Observatory, with all five sections displaying a different neuroimaging modality and/or visualisation type. Photo credit: Authors MW and DB.\n\nThis collaborative, high-resolution approach to visualisation has much to offer the neuroimaging community. Particularly:\n\n1) The ability to view images at full resolution without the need for interruptive actions such as zooming or panning through an image.\n\n2) The ability for multiple practitioners to share the same, high resolution, view of data for discussion in a collaborative environment.\n\n3) The ability to simultaneously show many views of the same or complementary data; large-scale visualisation allows complementary data to be shown simultaneously and accessed by a turn of the head, which enables easy comparison.\n\nThese benefits are of particular value to collaborative interdisciplinary groups exploring such multi-modal imaging studies. One case study under exploration at Imperial College involves Multiple Sclerosis.\n\n\nCase study: Multiple sclerosis (MS)\n\nMS is an autoimmune disease that affects more than 100,000 people in the United Kingdom alone (Mackenzie et al., 2014). It has a debilitating effect on various body functions including vision, motor and cognition; while there are treatment options available there is currently no known cure. MS assessments are made using objective clinical criteria, supplemented by findings of lesions in the central nervous system that are detectable on MRI scans over a period of time and space (Polman et al., 2011). T1 images, T2 images, and contrast-enhanced MRI using gadolinium are all useful techniques in this regard (Bakshi et al., 2008). Other diagnostic tests include assessment of visual function (as visual deterioration occurs in over 80% of patients) using a Low-Contrast Sloan Letters Chart (Baier et al., 2005). Optical Coherence Tomography (OCT; Petzold et al., 2010) or Visual Evoked Potentials (VEP; Schlaeger et al., 2014) can also provide measures of retinal integrity and central nerve damage, respectively. Finally, functional tests and questionnaires can register cognitive and functional impairments. There are currently no known specific blood or cerebrospinal fluid (CSF)-borne biomarkers for MS (Polman et al., 2011), so diagnosis depends on a combination of these measures, and the relationship between these tests and disease progression (particularly in a predictive sense, e.g. Neema et al., 2009) is an area of active research.\n\nSynthesising and visualising the results of these varied tests is a challenge that is being addressed by current work on the DO. Lesion volume change and brain volume change data from analysis of MRI images are a critical components for tracking disease progression. Images from each contrast type (T1, T2, gadolinium-enhanced) provide unique information along with complementary limitations. The ability to register and view all modalities simultaneously enables the viewer to crosscheck the same regions of interest across large screens without the current need to toggle between screens or windows.\n\nIn addition, tools can be built to replicate inputs across imaging modalities, and between image sessions. A tool that highlights a region of interest on one modality can automatically replicate the marking of the same region across other modalities, on other sections of the DO display. Similarly, a lesion may be marked in the baseline image and have that mark replicated in a registered image from a follow-up session. The ability to view changes in these images simultaneously in the context of data collected from other tests such as OCT, VEP, functional criteria or radiological reports enables viewing of disparate sources of information in tight context.\n\nThe large display-area provides space to fit a timeline that incorporates imaging data, clinical events, treatments, written reports, and clinical test results; this gives a unique visualisation of the cause and effects of disease progression, treatments, and relapse events in MS. The flexibility and size of the display space enables novel visualisations, such as the scope to concurrently view individual results from a group of research subjects, or to view multiple sets of longitudinal data from a single subject. Clinicians and researchers can view, correlate, and cross-validate findings across heterogeneous data types within a single environment. As it is designed to be a collaborative environment for exploring images, multiple clinicians can highlight and share findings from different modalities and sources.\n\nEnvironments such as the DO may one day become commonplace, however currently they are an expensive rarity, with only a few comparable systems existing worldwide (e.g. the ‘HIPerWall’ at University of California, San Diego). This currently strongly limits their accessibility to many researchers and clinicians. These constraints make the use of a tool such as the DO in current clinical practice impractical. The DO is more effectively deployed for clinical research purposes, or perhaps in consultant meetings, when high-level discussion of an individual case is required.\n\n\nTechnical considerations\n\nFrom a technical perspective, software used within such high-resolution environments must be adapted to cope with higher pixel densities and to work across a network of rendering computers. This is rarely a straightforward change, although with the advent of new rendering systems it is becoming easier. In general, vector-based graphics systems that display neuroimaging data as a 3D mesh using rendering engines like OpenGL (e.g. Surf Ice) work better than bitmap-based display tools, that are limited by the (often poor) underlying resolution of the images themselves. Ideally, display software needs to evolve to support distributed visualisation systems able to support display across a large rendering surface, and scalability to support high-resolution environments. Also important will be the development of appropriate algorithms to support decision-making, for instance to highlight areas of potential interest to clinicians for review. This method of focussing attention and insight will be a critical area of development in the near future, and will need to have extremely high levels of robustness and reliability, particularly if algorithms will eventually have some input into clinical decision-making. Machine-learning platforms such as Google’s Tensorflow (Abadi et al., 2016) are likely to be important components of such systems.\n\nOne potential area of investigation enabled by the DO is the quantification of how images are used within a visualisation space, particularly which data and which image regions are of most interest to clinicians. The key means to doing this is via head- and eye-tracking systems, which are starting to become available within such visualisation spaces. This would provide a means of identifying patterns of behaviour in how clinicians use images to identify the most salient features. One hypothesis worthy of further investigation is to explore how clinicians with different levels of experience explore, manipulate, and interpret a set of different images. Eye tracking can also be used to improve user experience, to ensure that the most commonly accessed information is placed in prominent display areas.\n\n\nConclusions\n\nVisualisation spaces such as the DO are relatively novel environments, and discovering the most effective ways of using them is still an on-going process. High-resolution spaces like the DO offer greater fidelity over previous large-scale systems, which can potentially drive greater insights. The large size of the space enables easy comparison and synthesis of multiple types of data, most obviously imaging formats, but also other clinical or research data types. Finally, the immersive collaboration space it provides can help to initiate and strengthen multi-disciplinary collaboration between clinicians, researchers, and data scientists. Large format displays like the DO have much to offer and will likely form an important part of future research and clinical practice.",
"appendix": "Author contributions\n\n\n\nAll authors contributed to the first draft of the manuscript, and also were involved in revisions and editing of the final version. All authors have agreed to the final content.\n\n\nCompeting interests\n\n\n\nMW’s primary employer is Imanova Ltd., a privately owned company specialising in contract research work for the pharmaceutical and bio-technology industries.\n\nMY and DB have no competing interests to declare.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAbadi M, Agarwal A, Barham P, et al.: Tensorflow: Large-scale machine learning on heterogeneous distributed systems. arXiv preprint arXiv: 1603.04467. 2016. Reference Source\n\nBailey DL, Pichler BJ, Gückel B, et al.: Combined PET/MRI: Multi-modality Multi-parametric Imaging Is Here: Summary Report of the 4th International Workshop on PET/MR Imaging; February 23–27, 2015, Tübingen, Germany. Mol Imaging Biol. 2015; 17(5): 595–608. PubMed Abstract | Publisher Full Text\n\nBaier ML, Cutter GR, Rudick RA, et al.: Low-contrast letter acuity testing captures visual dysfunction in patients with multiple sclerosis. Neurology. 2005; 64(6): 992–995. PubMed Abstract | Publisher Full Text\n\nCarhart-Harris RL, Muthukumaraswamy S, Roseman L, et al.: Neural correlates of the LSD experience revealed by multimodal neuroimaging. Proc Natl Acad Sci U S A. 2016; 113(17): 4853–4858. PubMed Abstract | Publisher Full Text | Free Full Text\n\nColasanti A, Guo Q, Giannetti P, et al.: Hippocampal Neuroinflammation, Functional Connectivity, and Depressive Symptoms in Multiple Sclerosis. Biol Psychiatry. 2016; 80(1): 62–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFebretti A, Nishimoto A, Thigpen T, et al.: CAVE2: a hybrid reality environment for immersive simulation and information analysis. In IS&T/SPIE Electronic Imaging. International Society for Optics and Photonics, 2013; 864903. Publisher Full Text\n\nGunn R, Coello C, Searle G: Molecular Imaging And Kinetic Analysis Toolbox (MIAKAT) - A Quantitative Software Package for the Analysis of PET Neuroimaging Data. J Nucl Med. 2016; 57(supplement 2): 1928. Reference Source\n\nHuster RJ, Debener S, Eichele T, et al.: Methods for simultaneous EEG-fMRI: an introductory review. J Neurosci. 2012; 32(18): 6053–6060. PubMed Abstract | Publisher Full Text\n\nKatal A, Wazid M, Goudar RH: Big data: issues, challenges, tools and good practices. In Contemporary Computing (IC3), 2013 Sixth International Conference on, IEEE, 2013; 404–409. Publisher Full Text\n\nMackenzie IS, Morant SV, Bloomfield GA, et al.: Incidence and prevalence of multiple sclerosis in the UK 1990–2010: a descriptive study in the General Practice Research Database. J Neurol Neurosurg Psychiatry. 2014; 85(1): 76–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMadan CR: Creating 3D visualizations of MRI data: A brief guide [version 1; referees: 3 approved]. F1000Res. 2015; 4: 466. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarx V: Biology: The big challenges of big data. Nature. 2013; 498(7453): 255–260. PubMed Abstract | Publisher Full Text\n\nMatthews PM, Rabiner I, Gunn R: Non-invasive imaging in experimental medicine for drug development. Curr Opin Pharmacol. 2011; 11(5): 501–507. PubMed Abstract | Publisher Full Text\n\nMoeller S, Yacoub E, Olman CA, et al.: Multiband multislice GE-EPI at 7 tesla, with 16-fold acceleration using partial parallel imaging with application to high spatial and temporal whole-brain fMRI. Magn Reson Med. 2010; 63(5): 1144–1153. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeema M, Arora A, Healy BC, et al.: Deep gray matter involvement on brain MRI scans is associated with clinical progression in multiple sclerosis. J Neuroimaging. 2009; 19(1): 3–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel AB, Birla M, Nair U: Addressing big data problem using Hadoop and Map Reduce. In 2012 Nirma University International Conference on Engineering (NUiCONE). IEEE, 2012; 1–5. Publisher Full Text\n\nPetzold A, de Boer JF, Schippling S, et al.: Optical coherence tomography in multiple sclerosis: a systematic review and meta-analysis. Lancet Neurol. 2010; 9(9): 921–932. PubMed Abstract | Publisher Full Text\n\nPolman CH, Reingold SC, Banwell B, et al.: Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011; 69(2): 292–302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRabiner EA, Beaver J, Makwana A, et al.: Pharmacological differentiation of opioid receptor antagonists by molecular and functional imaging of target occupancy and food reward-related brain activation in humans. Mol Psychiatry. 2011; 16(8): 826–835. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchlaeger R, Schindler C, Grize L, et al.: Combined visual and motor evoked potentials predict multiple sclerosis disability after 20 years. Mult Scler. 2014; 20(10): 1348–1354. PubMed Abstract | Publisher Full Text\n\nVan Essen DC, Smith SM, Barch DM, et al.: The WU-Minn Human Connectome Project: an overview. Neuroimage. 2013; 80: 62–79. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "16063",
"date": "08 Sep 2016",
"name": "Jens Foell",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes a novel way to display large-scale data, including (but not limited to) data with a clear relation to neuroinformatics, with the goal of increasing the utility of complex multi-modal data. This is a relevant topic for neuroinformatics and neuroimaging, as it ensures that data accessibility advances in the same pace as the possibilities of data acquisition, data analysis, and data storage do. This manuscript is well written and provides a clear and comprehensive overview of the capabilities and logistics of one particular display system which is suitable to investigate functional or structural neuroimaging data and to compare changes over time.\nFor these reasons, this manuscript will be relevant and interesting to the readership of F1000Research and I recommend its indexing after some minor technical points have been addressed (see below).\nMinor recommended changes:\n1) The very first citation (\"Katal et al., 2015\"; paragraph 1, line 4) does not contain a hyperlink. The same is true for the citation \"Bakshi et al., 2008\" in the first paragraph of the Case Study section.\n2) The abbreviation CAVE should be explained at its first instance in the text (5th paragraph)",
"responses": []
},
{
"id": "16066",
"date": "19 Sep 2016",
"name": "Joshua Balsters",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript Wall et al. introduce the future potential of large display systems for visualizing neuroimaging data, making particular reference to their relevance for clinical neuroimaging by visualising multiple sources of neuroimaging information in parallel. I found this manuscript to be timely and well written. The MS case study also provides a clear example of how these tools can be used in real life.\n\nI whole-heartedly recommend this article is indexed after a couple minor grammatical errors are addressed.\n\nThird paragraph of the introduction – I think you mean to say “MRI and MEG have also been ‘fused’ …” rather than ‘used’.\n\nPg 5 paragraph starting “Synthesising and visualising…” – change “are a critical components” to “are critical components”.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2157
|
https://f1000research.com/articles/5-275/v1
|
03 Mar 16
|
{
"type": "Research Note",
"title": "Illustrating and homology modeling the proteins of the Zika virus",
"authors": [
"Sean Ekins",
"John Liebler",
"Bruno J. Neves",
"Warren G. Lewis",
"Megan Coffee",
"Rachelle Bienstock",
"Christopher Southan",
"Carolina H. Andrade",
"John Liebler",
"Bruno J. Neves",
"Warren G. Lewis",
"Megan Coffee",
"Rachelle Bienstock",
"Christopher Southan",
"Carolina H. Andrade"
],
"abstract": "The Zika virus (ZIKV) is a flavivirus of the family Flaviviridae, which is similar to dengue virus, yellow fever and West Nile virus. Recent outbreaks in South America, Latin America, the Caribbean and in particular Brazil have led to concern for the spread of the disease and potential to cause Guillain-Barré syndrome and microcephaly. Although ZIKV has been known of for over 60 years there is very little in the way of knowledge of the virus with few publications and no crystal structures. No antivirals have been tested against it either in vitro or in vivo. ZIKV therefore epitomizes a neglected disease. Several suggested steps have been proposed which could be taken to initiate ZIKV antiviral drug discovery using both high throughput screens as well as structure-based design based on homology models for the key proteins. We now describe preliminary homology models created for NS5, FtsJ, NS4B, NS4A, HELICc, DEXDc, peptidase S7, NS2B, NS2A, NS1, E stem, glycoprotein M, propeptide, capsid and glycoprotein E using SWISS-MODEL. Eleven out of 15 models pass our criteria for selection. While a ZIKV glycoprotein E homology model was initially described in the immature conformation as a trimer, we now describe the mature dimer conformer which allowed the construction of an illustration of the complete virion. By comparing illustrations of ZIKV based on this new homology model and the dengue virus crystal structure we propose potential differences that could be exploited for antiviral and vaccine design. The prediction of sites for glycosylation on this protein may also be useful in this regard. While we await a cryo-EM structure of ZIKV and eventual crystal structures of the individual proteins, these homology models provide the community with a starting point for structure-based design of drugs and vaccines as well as a for computational virtual screening.",
"keywords": [
"Aedes mosquito",
"dengue virus",
"drug discovery",
"ebola virus",
"flavivirus",
"microcephaly",
"yellow fever",
"Zika virus"
],
"content": "Introduction\n\nAll flaviviruses are spherical and contain a genome of approximately 11kb that functions as mRNA and encodes a polyprotein that leads to 10 proteins1. Examples include dengue virus, yellow fever and West Nile virus2. The recent pandemic of ZIKV occurring in South America, Latin America, the Caribbean and in particular Brazil spread by the Aedes mosquito has awakened dormant interest in this flavivirus which is a mild dengue-like disease3. However several documented cases of Guillain-Barré syndrome and other neurologic conditions represent important complications of the disease. In recent weeks the extent of the disease has also become apparent as new discoveries and announcements are made almost daily. Though clearly we have a considerable number of significant gaps in our knowledge which need addressing4.\n\nThe most concerning issue however is microcephaly observed in women who had ZIKV during pregnancy. There have been multiple cases of ZIKV found in fetal or newborn brain tissue that had signs of prenatal damage. The virus seems to have neurotropism in fetal brains, which may account for the presumed association between the infection and microcephaly5,6. The fetus in the recent case study had microcephaly with calcifications and ZIKV was found in the brain6. The ZIKV strain was identified as from French Polynesia (GenBank accession number KJ776791) and several polymorphisms were noted in the NS1, NS4B and FtsJ like methyltransferase regions. While the findings are not absolute proof that ZIKV causes microcephaly, the evidence from this case report strengthens the linkage7. Experts involved in the decision on the World Health Organization determined Public Health Emergency of International Concern (PHEIC) recommended the need for more research into the microcephaly link and need for an animal model to be developed. This group also interestingly called for open data sharing8. Early work 45 years ago in inoculated newborn mice showed that ZIKV had neurological effects, enlarging astroglial cells and destroying pyriform cells. At the same time virus formation within the endoplasmic reticulum was also visualized9. We are not aware of any studies of effects of ZIKV on human brain or brain cells. Localization of such viruses to the brain is not unusual for flaviviruses i.e. West Nile virus and this tropism may arise from viral binding to glycosaminoglycans, as has been observed for dengue virus in human microvascular endothelial cells10. Heparan sulfate and the C-type lectin DC-SIGN (dendritic cell-specific intercellular adhesion molecule 3-grabbing nonintegrin) are well characterized attachment structures for flaviviruses on cells. Interfering with glycan binding is one potential approach to preventing virus entry. Another is to acidify the endosome as has been demonstrated in vitro with chloroquine for dengue virus infection2. Several entry and adhesion factors, including DC-SIGN, Tyro3, and AXL as well as others, have been shown to permit ZIKV entry in human skin cells11.\n\nThe routes for transmission of ZIKV besides mosquito are of some concern. Recent US CDC guidance to pregnant women describes precautions against sexual transmission of ZIKV12 and that the virus can persist for up to 12 weeks13. Possible ZIKV transmission through blood transfusion in French Polynesia was described by detecting the virus in 3% of asymptomatic blood donors14. Given how widespread ZIKV has become, there is a risk of depleting the blood supply, if donation after potential virus exposure is deferred. Methods have also been developed to inactivate ZIKV in plasma using amotosalen and UVA illumination15. There are issues with detection of ZIKV as a false positive dengue NS1 antigen test in a traveler to Switzerland was found to have the virus later. Therefore, cross-reactivity appears to be an issue in detection16 and this also suggests the need for better diagnostics to be developed.\n\nStructural knowledge of the ZIKV proteins may allow us to understand exposed epitopes which will facilitate the development of specific diagnostic reagents that differentiate it from dengue and other flaviviruses. Furthermore, open sharing of the three-dimensional arrangement of viral surface proteins could allow the mapping of potential neutralizing epitopes, guiding efforts to rationally design effective vaccines. We recently developed a preliminary model for ZIKV glycoprotein E based on the dengue virus glycoprotein E late stage fusion intermediate as a trimer4. We now provide homology models of the glycoprotein E based on a dimer structure as well as attempts at modeling the other proteins in ZIKV. We have also investigated the likely glycosylation sites of the ZIKV envelope glycoprotein. Glycosylation may obstruct the binding of antibodies, or block access to potential underlying peptide antigens, so glycans may be an important consideration in diagnostic and vaccine development. Previously, glycosylation analysis using several computational tools predicted mammalian N-linked glycosylation at Asn-154 in most ZIKV strains17, a site known to be important in other flaviviruses. N-glycosylation of the dengue envelope glycoprotein at two sites has been shown to mediate interactions with DC-SIGN18. Insect cell expression of dengue shows both high-mannose and complex glycans19. However, flavivirus glycosylation in model systems such as insect cell culture and mammalian tumor cell lines may not represent the true infective insect and mammalian glycoprotein. Mammalian O-linked glycosylation was predicted at Thr-245 and Thr-381 in some isolates17. Yet, reliable prediction is hampered by a lack of O-glycosylation consensus sequences and preferences for O-glycosylation driven by structural characteristics20.\n\nUnderstanding the three-dimensional structure of antigenic ZIKV proteins may help accelerate the development of antibodies for diagnostics and rationally designed vaccines. In addition, the comparison of the assembled surface glycoprotein of ZIKV with that of dengue virus may help understand the accessible epitopes for the development of anti-flaviral vaccines in general. There is considerable prior work including structure-based design and virtual screening for dengue, yellow fever and other related flaviviruses to develop antivirals to target envelope glycoproteins21–26, guanylyltransferase27, capsid protein NS3 helicase, NS2B-NS3 protease and NS5 polymerase28,29, as well as whole cell screens30 which have produced many molecules potentially useful against ZIKV in vitro. Early work has only tested a small number of FDA-approved drugs against ZIKV including (EC50 in parenthesis) interferon (34.3 IU/ml), ribavirin (143 ug/ml), 6-azauridine (1.5 ug/ml) and glycyrrhizin (384 ug/ml)31. A recent paper by Hamel et al., from 2015 also showed interferon inhibited ZIKV replication in primary skin fibroblasts11.\n\nHowever, the use of compounds against ZIKV should take into account the treatment of pregnant women, and many of the potential options are unsuitable for use in pregnancy because of toxicity and/or teratogenicity. Despite limited human data, the available data in animal models suggests caution. Azauridine is highly toxic to the fetus in model systems (for example 32). Ribavirin is not recommended for use in pregnancy due to embryotoxic and teratogenic effects33. Interferon is a potential abortifacient34. We also need to consider the treatment of fetus as well as children (that might become infected after birth) and the relatively small subsection of FDA approved drugs that are approved for pediatric use35. Therefore, alternative potential drugs for ZIKV are needed. The risks of medication use in pregnancy are notable. In particular, these include teratogenicity concerns. There is also the issue that the disease does not usually pose a direct risk to pregnant women themselves, so it’s important any drug, which will not be improving their own health, does not damage their health. Pregnancy can also create increased risks and liver problems, a particular concern with any new drug as it can also affect drug distribution. In pregnancy, a ZIKV infection at 13 weeks gestation was coupled with persistent virus in a fetus at 32 weeks6. Treatment of symptomatic pregnant women may reduce risks of transmission to the fetus. A potential drug for ZIKV could also protect fetuses from damage by reducing transmission in the general population. If the drug appeared to reduce the duration of symptoms (which though mild can be annoying) and in turn reduced viral load, reducing the chance of transmission, this could benefit them. For example, cholera patients are often given antibiotics to reduce transmission. Also those with the flu are prescribed Tamiflu/Oseltamivir to reduce the duration of symptoms, minimize the severity of symptoms, but its use might also reduce transmission of flu in the general population. Since ZIKV has been found in semen many weeks after symptoms resolve12, treatment of male partners may reduce viral load and reduce the long term risk of transmission.\n\nTo help accelerate drug discovery through computational analysis, we have now developed homology models of ZIKV proteins that may serve as potential drug and vaccine targets. To complement high-throughput screening efforts, we could perform virtual screening against the proteins in ZIKV. While there are crystal structures for proteins from dengue36,37, yellow fever, West Nile virus and other flaviviruses38–44 there are (to date) none for ZIKV. Therefore we are limited to generating homology models, although the close evolutionary relationships between flavivirus and their component proteins and genomes represents a valid approach45.\n\n\nMethods\n\nAs a prelude to modeling we assessed what 3D structural data had significant identity scores to a representative ZIKV polyprotein. For this we chose UniProtKB Q32ZE1_9FLAV. While we have requested the promotion of this entry to the Swiss-Prot expert review level it has been selected as the representative sequence for the UniRef90_Q32ZE1 entry that currently clusters 108 ZIKV individual sequence entries at 90% (or above) amino acid identity. We then performed a BLAST search of this against Protein Data Bank (PDB) sequence entries. Protein BLAST analysis was also performed for each ZIKV protein sequence46 to identify the closest proteins47 and understand potential evolution.\n\nThe amino acid sequences of ZIKV strain (GenBank accession number KJ77679148) were retrieved from the GenBank database49 and used as targets for homology modelling using the SWISS-MODEL server50,51. The latter performed the target-template sequence alignment after searching the putative X-ray template proteins in PDB for generating the 3D models for all target sequences. The best homology models were selected according to Global Model Quality Estimation (GMQE) and QMEAN statistical parameters. GMQE is a quality estimation which combines properties from the target-template alignment. The quality estimate ranges between 0 and 1 with higher values for better models. QMEAN4 scoring function consisting of a linear combination of four structural descriptors as described elsewhere in more detail52,53. The pseudo energies returned from the four descriptors are related to what we would expect from high resolution X-ray structures of similar size using a Z-score scheme. Further, built models were exported to the SAVES server Version 454 and their overall stereochemical quality, including backbone torsional angles through the Ramachandran plot, was checked according to PROCHECK55. Lastly, each model was refined by an energy minimization protocol, using the Smart Minimizer algorithm in Discovery Studio version 4.1 (Biovia, San Diego, CA).\n\nMammalian N-glycosylation sites were predicted for glycoprotein E by submitting the sequence to web-based tools namely N-GlycoSite56, GlycoEP57,58 and NetNGlyc Version 1.059.\n\nThe Zika virion illustrations were created by combining the homology model of the envelope ZIKV glycoprotein E with the symmetry data from the dengue virus envelope. PDB ID:1K4R60 contains the coordinates for three copies of the protein subunit of the dengue virus envelope, along with the symmetry data necessary to create the 180-subunit icosahedral structure of the complete viral envelope. The PyMOL Molecular Graphics System, Version 1.7.6.0. Schrödinger, LLC. was used to export the surface models of the three proteins in .obj format. Then they were imported into Lightwave 3D (NewTek, San Antonio, TX) where the symmetry data was used to instance copies of the model into the icosahedral envelope. The entire structure was copied several times and lighting applied as a surfacing effect to create a visually pleasing composition, and the image rendered out.\n\nThe next step was to import into Pymol the homology model of the ZIKV envelope protein which was homology modeled using PDB ID: 3P54 (from Japanese Encephalitis Virus) as a template. A surface model of this protein was exported from Pymol as an .obj, and imported into Lightwave in place of the dengue model, using the same symmetry operators to create the envelope array. Everything else about the picture was left the same (color, composition, lighting, etc...) to allow the structural differences to be more apparent, and that image rendered out as well.\n\nThe last step was to overlay the detailed area of the two images and create an animated gif to flip back and forth between the two images of ZIKV and dengue, again to allow the differences to be more clearly seen. The structure of the Zika virion could be explored in a similar manner, using known data from other flaviviruses as a guide.\n\nThe immature4 and mature (this study) homology models for glycoprotein E were compared using the ‘align and superimpose’ proteins protocol in Discovery Studio Version 4.1 (Biovia, San Diego, CA).\n\n\nResults\n\nA BLAST search of ZIKV polyprotein against PDB sequence entries shows the highest scoring matches with 55–70% sequence identity (Supplementary material S1). Protein BLAST analysis of the individual ZIKV protein sequences show that many of the proteins are similar to the same protein from Spondweni virus in 12 out of 15 cases (Table 1). Exceptions were: FtsJ which was closer to the Murray Valley encephalitis virus, NS1 which was more similar to dengue virus 3, and glycoprotein E which was closest to the dengue virus 1 protein. These results are in general accordance with whole sequence analyses45.\n\nThe SWISS-MODEL server was used to generate alignments (Supplementary material S2) and homology models for all Zika proteins (Table 2, Figure 1, Supplementary material S3). First, we selected suitable template protein structures in PDB, observing the following criteria: the template should have a high coverage (i.e., > 65% of target aligned to template) and sequence identity >30%. Then, we used GMQE and QMEAN4 scoring function as an initial criteria to discriminate good from bad models. Acceptable alignment values and higher GMQE and QMEAN4 scores were obtained during modeling, suggesting statistically acceptable homology models were generated for 11 proteins: NS5, FtsJ, HELICc, DEXDc, peptidase S7, NS1, E stem, glycoprotein M, propeptide, capsid, and glycoprotein E (Table 2, Figure 1 and Figure 2). The Ramachandran plots for these 11 proteins provide further evidence of their acceptability (Figure 2). On the other hand, because of low GMQE scores and of low coverage observed in X-ray template proteins available in the PDB, homology models for NS4B, NS4A, NS2B, and NS2A proteins appeared to have limitations regarding active sites and epitopes and they could not be validated.\n\nThe global and per-residue model quality has been assessed using the QMEAN scoring function53. For improved performance, weights of the individual QMEAN terms have been trained specifically for SWISS-MODEL50,51,71–73. GMQE = Global Model Quality Estimation, QMEAN4 is a scoring function consisting of a linear combination of four structural descriptors as described elsewhere in more detail52,53.\n\nSelected ZIKV NS5 (A), FtsJ (B), HELICc (C), DEXDc (D), Peptidase S7 (E), NS1 (F), E Stem (G), Glycoprotein M (H), Propeptide (I), Capsid (J), and Glycoprotein E (K) homology models (minimized proteins) that had good sequence coverage with template proteins developed with SWISS-MODEL.\n\nRamachandran plots for ZIKV NS5 (A), FtsJ (B), HELICc (C), DEXDc (D), Peptidase S7 (E), NS1 (F), E Stem (G), Glycoprotein M (H), Propeptide (I), Capsid (J), and Glycoprotein E (K) obtained by PROCHECK, showing the dihedral angles Psi and Phi of amino acid residues. Red represents most favored regions; yellow represents additional allowed regions; beige represents generously allowed regions; and white areas are disallowed regions.\n\nThe best NS5 homology model was built using the full-length Japanese encephalitis virus NS5 as a template (PDB ID: 4K6M)61. The homology model generated for the FtsJ protein was built using the crystal structure of the West Nile virus methyltransferase (PDB ID: 2OY0)62. The best HELICc model was built using dengue virus helicase/nucleoside triphosphatase catalytic domain (PDB ID: 2BHR)63 whereas the DEXDc model was built using the structure of the Murray Valley encephalitis virus RNA helicase (PDB ID: 2V8O)64. The peptidase S7 model was built using the West Nile virus Ns2B-Ns3 protease (PDB ID: 2YOL)65 and the NS1 model was built using West Nile virus non-structural protein 1 as template (PDB ID: 4O6D)66. In addition, the E stem model was built using the cryo-electron microscopy (cryo-EM) structure of dengue virus capsid protein heterotetramer (PDB ID: 3J2P)67 whereas the glycoprotein M model was built using the cryo-EM structure of dengue virus as a template (PDB ID: 3J27)67. The propeptide model was built using the crystal structure of the precursor membrane protein-envelope protein heterodimer from the dengue 2 virus at low pH (PDB ID: 3C5X)68 and the capsid model was built using the core (C) protein from West Nile virus, subtype Kunjin (PDB ID: 1SFK)69. Finally, the best glycoprotein E model for the mature protein was built using the following sequence taken from the polyprotein, where the part corresponding to E is from residues 291-592, while the IG-like domain III is from residues 601-693. The model used the crystal structure of the Japanese encephalitis virus envelope protein, strain SA-14-14-2 as a template (PDB ID: 3P54)70. After building of homology models, we performed an additional validation in order to explore stereochemical quality of dihedral angles phi against psi of amino acid residues in modeled structures and identify sterically allowed regions for these angles using PROCHECK analysis. The results shown in Table 2 and Figure 2 reveal that 58.4─70.3% residues of the modeled proteins are within the most favored regions (red), 27.1─43.3% residues of modeled proteins are within the additional allowed regions (yellow), 1.5─6.3% residues of modeled proteins are within the generously allowed regions (beige), and only 0.0─6.1% residues of modeled proteins are within the disallowed regions (white). These results showed that the overall stereochemical properties of the generated models were highly reliable and the models could be useful to future molecular modeling studies.\n\nSeveral web-based tools were used for N-glycosylation site predictions as it provides a more thorough approach. N-GlycoSite56 suggested N154 as a single N-glycosylation site matching the N-X-S/T/C consensus sequence. The same site was identified by GlycoEP using BPP settings (binary profile of patterns)57,58 giving a score of 0.65/1.00. NetNGlyc59 also gave the same predicted site, with a jury agreement of 6/9.\n\nA qualitative analysis of the Zika virion (which was constructed based on the dengue virion) can be compared to the dengue cryo-EM virion (Figure 3) and indicates that Zika appears to have slightly more raised ‘pimples’ on the surface. The glycoprotein E dimer in ZIKV also has a narrow ‘letter-box’ groove while the dengue virion has a bigger ‘pore‘ between the intersection of 5 dimers (5 fold axis). These differences are considerably more apparent in the animation (Supplementary material S4). It is important to note that the differences may also be artefacts of the homology modeling approach and template used for modeling ZIKV glycoprotein E.\n\nThe homology models developed using two different templates namely the immature protein which was based on the dengue crystal structure 4gsx as a template50,71–73 and the mature protein which was based on PDB ID:3P54 from Japanese encephalitis virus showed a large difference (RMSD 13.47Å) (Figure 4). These proteins also demonstrate differences around the pocket used centered on the residues 270-277.\n\n\nDiscussion\n\nThe genus Flavivirus consists of 70 viruses many of which can cause severe human disease. There have been few sequence analyses of ZIKV previously in comparison to other flaviviruses. The genus Flavivirus produces a monophyletic tree with ZIKV being closest to Spondweni virus74 while mosquito borne, tick borne and no-vector viruses cluster separately45. A BLAST analysis of all the ZIKV proteins in this study suggests for 12 of 15, their closest protein is in Spondweni virus (Table 1). More often strain sequences are compared within ZIKV and these showed variations in the NS5 gene75 and glycoprotein E17. This is important as it would suggest perhaps targeting other proteins would have less issue with resistance or variability due to the strain of ZIKV.\n\nIf we are to address ZIKV in the short term while we await a vaccine we need to rapidly identify an antiviral, and preferably one that can be used against other related flaviviruses. Ideally we would need to treat pregnant women and provide them with prophylaxis that was safe to them and their fetus. Such an antiviral could also be used to reduce transmission in the population in general (by reducing viral load and symptoms and/or duration). As noted a decade ago and is still is true today, no antiviral drug is approved for any flavivirus to date76. It has been suggested that one of the ways to target these viruses is to interfere with the NS2B/NS3 protease complex76. Understanding of flavivirus proteins and other RNA viruses has benefited from the EU funded project VIZIER77, in particular several West Nile virus, dengue virus and other flavivirus structures of NS3 or NS5 were solved during this project and allosteric inhibitor sites were identified on NS578. Multiple pharmaceutical companies have worked on this target for HCV leading to clinical candidates like IDX32079, danoprevir (ITMN-191/R7227)80, GS-925681 and others82,83. The only HCV protease targeting FDA approved drug is simeprevir, TMC43584,85 and its use is avoided in pregnancy. Other HCV protease compounds are in clinical trials or submitted for FDA approval including Ledipasvir (formerly GS-5885)86. Testing these molecules against ZIKV in vitro would be useful.\n\nWe recently described 6 steps which could be taken to kick start research on ZIKV4, one of which was to develop homology models for ZIKV proteins that are similar to those targeted by molecules that are also active against the dengue virus. Such an approach would then enable docking of compound libraries of known antivirals, FDA approved drugs or other compounds4. Ideally generating homology models with a single tool may not be enough. In particular, for those proteins with low sequence identity the use of servers and methods that use threading may be worthwhile (e.g. I-TASSER87–89). However these methods are generally only accessible to academics while others are required to license the technologies. This is ironic as these technologies were developed in most cases with NIH and NSF funds. An alternative commercial homology modeling approach (MODELLER) was also used and generated a NS5 homology model and the top hit was also the Japanese encephalitis virus RdRp domain (PDB ID: 4HDH) compared with PDB ID: 4K6M from SWISS-MODEL61. 4HDH also includes the ATP and zinc metal where the catalytic centers are. The dengue virus 3 polymerase (PDB ID: 4HHJ)90 has very high sequence homology and comes up as a potential target in MODELLER, which illustrates that all these viral RNA dependent polymerases are very similar.\n\nWhile it is likely that the eventual availability of crystal structures of ZIKV proteins would improve the results of docking, the homology models described here (Figure 1, Figure 2, Supplementary material S2) represent a starting point that can be used to help prioritize compounds for testing as described previously4. Proteins with templates above 25–40% sequence identity might suggest the proteins are related while below this is a twilight zone. Homology modeling is thought to fill in the gaps between proteins with x-ray structures and those with none91. Experimental testing of homology models and crystal structures indicate that a similar enrichment rate can be achieved when identifying active compounds in a set decoys92. Others have also described homology models that may be an excellent alternative when crystal structures are unavailable for human GPCRs93,94, and have led to the first identification of inhibitors of the Mycobacterium tuberculosis Topoisomerase I after virtual screening95,96 prior to the crystal structure becoming available97. Certainly there are still considerable challenges using homology models such as prediction of the correct binding pose98 but there are plenty of success stories98–100. While databases of homology models exist like MODBASE101 and SWISS-MODEL50,51,71–73 neither of these have any ZIKV protein homology models at the time of writing. There are many structural genomics initiatives and yet it would seem there are few if any continuing the work of VIZIER working on flaviviruses or emerging viruses.\n\nAvailability of structures are important as the structure of the ZIKV glycoprotein could be useful for design of antibodies selective for the virus which will be critical for the development of diagnostics, and understanding antibody binding also for the use of IV immunoglobulin in pregnancy and the organization of the epitopes on viral proteins may facilitate early work in vaccine development. There are further implications for understanding the antibody binding epitopes, which are sometimes shared between different flaviviruses. Broadly protective vaccines for flaviviruses may allow the simultaneous targeting of ZIKV and related viruses such as dengue102. Understanding glycosylation is therefore important. To date Asn-154 is mentioned in Faye et al.,17 as a glycosylation site, as are Thr-170 (mucin type O-linked glycosylation) and other mucin sites at Thr-245 and Thr-381. Other probably O-GlcNAC attachment sites (Ser-142, Ser-227, Thr-231, Ser-304, Thr-366, and Thr-381) were also predicted. Our analysis of N-glycosylation with 3 different websites suggest Asn-154 also as a likely site of N-glycosylation in agreement with dengue virus18.\n\nCryo-EM has been used to show how glycoprotein E dimers arrange on the surface of virions for flaviviruses including tick-borne encephalitis virus103, West Nile virus104, dengue virus 1105 and dengue virus 4106 (Supplementary material S3). The early work on the tick-borne encephalitis virus103 suggested 30 dimers on the surface and also pointed to how the glycoprotein E dimers can reorganize under low pH to form a trimer. The packing of the dimers in the dengue virion is different to tick-borne encephalitis with the glycoprotein E dimers showing 30 ‘herringbone rafts’ each containing three dimers to result in 180 copies of the protein67,107. West Nile virus again has a different arrangement with 60 trimers shown in the structure of the immature virus104 (Supplementary material S4). Even between the dengue serotypes 1, 2 and 4 for which there are cryo-EM structures105,106 it is apparent while the rafts are very similar (as are the sequence identities [60%]) there is a different charge distribution of the surface of each. Dengue serotype 2 had larger continuous patches of positive charges which was proposed to enable improved binding to heparan sulfate. This might also be the case for ZIKV in that the charge pattern is again different and could be key for vaccine development. The availability of virion structures makes it feasible to understand structure function of the complete virus such as assessment of membrane curvature and how organization of membrane proteins affects this43.\n\nA model of the Zika virion was constructed as an illustration using the homology model of the glycoprotein E dimer (Figure 3). While the combined protein sequence of glycoprotein E and the immunoglobulin like domain is closest to dengue virus 1 (57 percent identity, Table 1) the closest template was for the crystal structure of the Japanese encephalitis virus envelope protein (53.12 percent identity, Table 2). This would suggest the virion should more closely resemble that of dengue virus 1, while producing a homology model based on a more distant virus might not be ideal. The homology model of glycoprotein E developed for the mature conformation in this study is significantly different from that developed previously for the immature conformation (Figure 4). The proposed binding site centered around residues 270-277 appears shallower in the mature conformation and this would certainly affect the kinds of molecules that it could interact with. It might also point to the need to interfere with the immature conformation as preferable versus the mature conformation. Ultimately perhaps this model of the Zika virion could help us understand how drugs could access the virus. Viruses affecting pregnancy, like say Varicella which causes microcephaly and other developmental problems106,107, are often treated with IV immunoglobulin, i.e. antibodies, as well as antivirals to reduce the effect of the virus (or to avoid infection if given soon after exposure). The models could help us design combination approaches possibly targeting multiple proteins that might prevent drug resistance from occurring also.\n\nDoes having the homology models and the virion illustration help understand function? Well, the surface charge pattern might be inferred from the homology model and could be compared with dengue and other filoviruses for which there are cryo-EM structures. This may in turn present opportunities for vaccine design by indicating accessible surfaces and properties, allowing mapping of epitopes, design of accessible fragments and peptides for vaccine/diagnostic design. Vaccines themselves might be the only way to avoid the inevitable, otherwise, simply reducing the spread of ZIKV would just delay it. Women ultimately may just want to ‘get it over with’ and have ZIKV before they get pregnant and hope there is lasting immunity.\n\nIn summary, in the absence of crystal structures for any of the proteins comprising the ZIKV, we are left to attempt to construct homology models which we have done using the freely available SWISS-MODEL server. Further preparation of these models required freely available and commercial tools. In the case of the ZIKV glycoprotein E homology model, this has the added benefit of enabling the construction of a full virion. By comparing the Zika virion to the existing structures for other flaviviruses we can see similarities and differences on the surface (Supplementary material S4, Supplementary material S5). This relatively crude approach could help to understand how we might develop antivirals and vaccines against it. In addition we now provide homology models as a starting point for (small and large scale) docking studies and further evaluation which may complement other modeling efforts for ZIKV110. Ultimately the results of their use can be compared with using ZIKV crystal structures once generated.",
"appendix": "Author contributions\n\n\n\nAll authors contributed to the collaborative writing of this project. SE conceived and designed the experiments. SE, JL, BJN, WGL, CS and CHA carried out the research.\n\n\nCompeting interests\n\n\n\nS.E. works for Collaborations in Chemistry, Collaborations Pharmaceuticals, Inc. and Collaborative Drug Discovery, Inc.\n\n\nGrant information\n\nCS was supported by Wellcome Trust Grant (to the IUPHAR/BPS Guide to PHARMACOLOGY) Number 099156/Z/12/Z.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nDr’s Priscilla Yang, Andrew Marsh, Derek Gatherer, Lucio Freitas-Junior, Daniel Mietchen, Joel S Freundlich, Jair Siqueira-Neto, Antony J. Williams, Alex Perryman and Mr. Tom Stratton are thanked for their helpful discussions and tweets. Biovia is kindly acknowledged for providing Discovery Studio to SE.\n\n\nSupplementary material\n\nSupplementary material S1 (Alignments of ZIKV to other proteins), S2 (Alignment of individual ZIKV proteins for homology models), S4 (ZIKV versus dengue virion animation), and S5 (Published flavivirus cryo-EM structures [not to scale]).\n\nSupplementary material S3. PDB files for ZIKV homology models.\n\n\nReferences\n\nMlera L, Melik W, Bloom ME: The role of viral persistence in flavivirus biology. Pathog Dis. 2014; 71(2): 137–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPierson TC, Kielian M: Flaviviruses: braking the entering. Curr Opin Virol. 2013; 3(1): 3–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFauci AS, Morens DM: Zika Virus in the Americas--Yet Another Arbovirus Threat. N Engl J Med. 2016; 374(7): 601–4. PubMed Abstract | Publisher Full Text\n\nEkins S, Mietchen D, Coffee M, et al.: Open drug discovery for the Zika virus [version 1; referees: awaiting peer review]. F1000Res. 2016; 5: 150. Publisher Full Text\n\nMartines RB, Bhatnagar J, Keating MK, et al.: Notes from the Field: Evidence of Zika Virus Infection in Brain and Placental Tissues from Two Congenitally Infected Newborns and Two Fetal Losses - Brazil, 2015. MMWR Morb Mortal Wkly Rep. 2016; 65(6): 159–60. PubMed Abstract | Publisher Full Text\n\nMlakar J, Korva M, Tul N, et al.: Zika Virus Associated with Microcephaly. N Engl J Med. 2016. PubMed Abstract | Publisher Full Text\n\nRubin EJ, Greene MF, Baden LR: Zika Virus and Microcephaly. N Engl J Med. 2016. PubMed Abstract | Publisher Full Text\n\nHeymann DL, Hodgson A, Sall AA, et al.: Zika virus and microcephaly: why is this situation a PHEIC? Lancet. 2016; 387(10020): 719–21. PubMed Abstract | Publisher Full Text\n\nBell TM, Field EJ, Narang HK: Zika virus infection of the central nervous system of mice. Arch Gesamte Virusforsch. 1971; 35(2): 183–93. PubMed Abstract | Publisher Full Text\n\nVervaeke P, Alen M, Noppen S, et al.: Sulfated Escherichia coli K5 polysaccharide derivatives inhibit dengue virus infection of human microvascular endothelial cells by interacting with the viral envelope protein E domain III. PLoS One. 2013; 8(8): e74035. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamel R, Dejarnac O, Wichit S, et al.: Biology of Zika Virus Infection in Human Skin Cells. J Virol. 2015; 89(17): 8880–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnon. Zika and Sexual Transmission. 2016. Reference Source\n\nOduyebo T, Petersen EE, Rasmussen SA, et al.: Update: Interim Guidelines for Health Care Providers Caring for Pregnant Women and Women of Reproductive Age with Possible Zika Virus Exposure - United States, 2016. MMWR Morb Mortal Wkly Rep. 2016; 65(5): 122–7. PubMed Abstract | Publisher Full Text\n\nMusso D, Nhan T, Robin E, et al.: Potential for Zika virus transmission through blood transfusion demonstrated during an outbreak in French Polynesia, November 2013 to February 2014. Euro Surveill. 2014; 19(14). pii: 20761. PubMed Abstract | Publisher Full Text\n\nAubry M, Richard V, Green J, et al.: Inactivation of Zika virus in plasma with amotosalen and ultraviolet A illumination. Transfusion. 2016; 56(1): 33–40. PubMed Abstract | Publisher Full Text\n\nGyurech D, Schilling J, Schmidt-Chanasit J, et al.: False positive dengue NS1 antigen test in a traveller with an acute Zika virus infection imported into Switzerland. Swiss Med Wkly. 2016; 146: w14296. PubMed Abstract\n\nFaye O, Freire CC, Iamarino A, et al.: Molecular evolution of Zika virus during its emergence in the 20th century. PLoS Negl Trop Dis. 2014; 8(1): e2636. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlen MM, Kaptein SJ, De Burghgraeve T, et al.: Antiviral activity of carbohydrate-binding agents and the role of DC-SIGN in dengue virus infection. Virology. 2009; 387(1): 67–75. PubMed Abstract | Publisher Full Text\n\nLei Y, Yu H, Dong Y, et al.: Characterization of N-Glycan Structures on the Surface of Mature Dengue 2 Virus Derived from Insect Cells. PLoS One. 2015; 10(7): e0132122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNishikawa I, Nakajima Y, Ito M, et al.: Computational prediction of O-linked glycosylation sites that preferentially map on intrinsically disordered regions of extracellular proteins. Int J Mol Sci. 2010; 11(12): 4991–5008. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Z, Khaliq M, Zhou Z, et al.: Design, synthesis, and biological evaluation of antiviral agents targeting flavivirus envelope proteins. J Med Chem. 2008; 51(15): 4660–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoh MK, Yip A, Zhang S, et al.: A small molecule fusion inhibitor of dengue virus. Antiviral Res. 2009; 84(3): 260–6. PubMed Abstract | Publisher Full Text\n\nWang QY, Patel SJ, Vangrevelinghe E, et al.: A small-molecule dengue virus entry inhibitor. Antimicrob Agents Chemother. 2009; 53(5): 1823–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMayhoub AS, Khaliq M, Kuhn RJ, et al.: Design, synthesis, and biological evaluation of thiazoles targeting flavivirus envelope proteins. J Med Chem. 2011; 54(6): 1704–14. PubMed Abstract | Publisher Full Text\n\nSchmidt AG, Lee K, Yang PL, et al.: Small-molecule inhibitors of dengue-virus entry. PLoS Pathog. 2012; 8(4): e1002627. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Z, Khaliq M, Suk JE, et al.: Antiviral compounds discovered by virtual screening of small-molecule libraries against dengue virus E protein. ACS Chem Biol. 2008; 3(12): 765–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStahla-Beek HJ, April DG, Saeedi BJ, et al.: Identification of a novel antiviral inhibitor of the flavivirus guanylyltransferase enzyme. J Virol. 2012; 86(16): 8730–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBehnam MA, Nitsche C, Boldescu V, et al.: The Medicinal Chemistry of Dengue Virus. J Med Chem. 2016. PubMed Abstract | Publisher Full Text\n\nVincetti P, Caporuscio F, Kaptein S, et al.: Discovery of Multitarget Antivirals Acting on Both the Dengue Virus NS5-NS3 Interaction and the Host Src/Fyn Kinases. J Med Chem. 2015; 58(12): 4964–75. PubMed Abstract | Publisher Full Text\n\nShum D, Smith JL, Hirsch AJ, et al.: High-content assay to identify inhibitors of dengue virus infection. Assay Drug Dev Technol. 2010; 8(5): 553–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrance JM, Scaramozzino N, Jouan A, et al.: Interferon, ribavirin, 6-azauridine and glycyrrhizin: antiviral compounds active against pathogenic flaviviruses. Antiviral Res. 2003; 58(1): 73–9. PubMed Abstract | Publisher Full Text\n\nRaska K Jr, Zedeck MS, Welch AD: Relationship between the metabolic effects and the pregnancy-interrupting property of 6-azauridine in mice. Biochem Pharmacol. 1966; 15(12): 2136–8. PubMed Abstract | Publisher Full Text\n\nKochhar DM, Penner JD, Knudsen TB: Embryotoxic, teratogenic, and metabolic effects of ribavirin in mice. Toxicol Appl Pharmacol. 1980; 52(1): 99–112. PubMed Abstract | Publisher Full Text\n\nEntrican G: Immune regulation during pregnancy and host-pathogen interactions in infectious abortion. J Comp Pathol. 2002; 126(2–3): 79–94. PubMed Abstract | Publisher Full Text\n\nBlatt J, Farag S, Corey SJ, et al.: Expanding the scope of drug repurposing in pediatrics: the Children's Pharmacy Collaborative. Drug Discov Today. 2014; 19(11): 1696–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlein DE, Choi JL, Harrison SC: Structure of a dengue virus envelope protein late-stage fusion intermediate. J Virol. 2013; 87(4): 2287–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nModis Y, Ogata S, Clements D, et al.: A ligand-binding pocket in the dengue virus envelope glycoprotein. Proc Natl Acad Sci U S A. 2003; 100(12): 6986–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu J, Bera AK, Kuhn RJ, et al.: Structure of the Flavivirus helicase: implications for catalytic activity, protein interactions, and proteolytic processing. J Virol. 2005; 79(16): 10268–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeiss BJ, Thompson AA, Andrews AJ, et al.: Analysis of flavivirus NS5 methyltransferase cap binding. J Mol Biol. 2009; 385(5): 1643–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Y, Corver J, Chipman PR, et al.: Structures of immature flavivirus particles. EMBO J. 2003; 22(11): 2604–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrecher M, Chen H, Li Z, et al.: Identification and Characterization of Novel Broad-Spectrum Inhibitors of the Flavivirus Methyltransferase. ACS Infect Dis. 2015; 1(8): 340–349. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkey DL, Brown WC, Konwerski JR, et al.: Use of massively multiple merged data for low-resolution S-SAD phasing and refinement of flavivirus NS1. Acta Crystallogr D Biol Crystallogr. 2014; 70(Pt 10): 2719–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang W, Kaufmann B, Chipman PR, et al.: Membrane curvature in flaviviruses. J Struct Biol. 2013; 183(1): 86–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAleshin AE, Shiryaev SA, Strongin AY, et al.: Structural evidence for regulation and specificity of flaviviral proteases and evolution of the Flaviviridae fold. Protein Sci. 2007; 16(5): 795–806. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuno G, Chang GJ, Tsuchiya KR, et al.: Phylogeny of the genus Flavivirus. J Virol. 1998; 72(1): 73–83. PubMed Abstract | Free Full Text\n\nAnon. BLAST. Reference Source\n\nAltschul SF, Gish W, Miller W, et al.: Basic local alignment search tool. J Mol Biol. 1990; 215(3): 403–10. PubMed Abstract | Publisher Full Text\n\nBaronti C, Piorkowski G, Charrel RN, et al.: Complete coding sequence of Zika virus from a French polynesia outbreak in 2013. Genome Announc. 2014; 2(3): pii: e00500-14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenson DA, Cavanaugh M, Clark K, et al.: GenBank. Nucleic Acids Res. 2013; 41(Database issue): D36–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBiasini M, Bienert S, Waterhouse A, et al.: SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information. Nucleic Acids Res. 2014; 42(Web Server issue): W252–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBordoli L, Kiefer F, Arnold K, et al.: Protein structure homology modeling using SWISS-MODEL workspace. Nat Protoc. 2009; 4(1): 1–13. PubMed Abstract | Publisher Full Text\n\nBenkert P, Tosatto SC, Schomburg D: QMEAN: A comprehensive scoring function for model quality assessment. Proteins. 2008; 71(1): 261–77. PubMed Abstract | Publisher Full Text\n\nBenkert P, Künzli M, Schwede T: QMEAN server for protein model quality estimation. Nucleic Acids Res. 2009; 37(Web Server issue): W510–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnon. The Structure Analysis and Verification Server. Reference Source\n\nAnon. PROCHECK and PROCHECK-NMR. Reference Source\n\nAnon. N-GlycoSite. Reference Source\n\nAnon. GlycoEP. Reference Source\n\nChauhan JS, Rao A, Raghava GP: In silico platform for prediction of N-, O- and C-glycosites in eukaryotic protein sequences. PLoS One. 2013; 8(6): e67008. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnon. NetNGlyc. Reference Source\n\nKuhn RJ, Zhang W, Rossmann MG, et al.: Structure of dengue virus: implications for flavivirus organization, maturation, and fusion. Cell. 2002; 108(5): 717–725. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu G, Gong P: Crystal Structure of the full-length Japanese encephalitis virus NS5 reveals a conserved methyltransferase-polymerase interface. PLoS Pathog. 2013; 9(8): e1003549. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Y, Ray D, Zhao Y, et al.: Structure and function of flavivirus NS5 methyltransferase. J Virol. 2007; 81(8): 3891–903. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu T, Sampath A, Chao A, et al.: Structure of the Dengue virus helicase/nucleoside triphosphatase catalytic domain at a resolution of 2.4 Å. J Virol. 2005; 79(16): 10278–88. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMancini EJ, Assenberg R, Verma A, et al.: Structure of the Murray Valley encephalitis virus RNA helicase at 1.9 Angstrom resolution. Protein Sci. 2007; 16(10): 2294–300. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHammamy MZ, Haase C, Hammami M, et al.: Development and characterization of new peptidomimetic inhibitors of the West Nile virus NS2B-NS3 protease. ChemMedChem. 2013; 8(2): 231–41. PubMed Abstract | Publisher Full Text\n\nAkey DL, Brown WC, Dutta S, et al.: Flavivirus NS1 structures reveal surfaces for associations with membranes and the immune system. Science. 2014; 343(6173): 881–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang X, Ge P, Yu X, et al.: Cryo-EM structure of the mature dengue virus at 3.5-Å resolution. Nat Struct Mol Biol. 2013; 20(1): 105–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi L, Lok SM, Yu IM, et al.: The flavivirus precursor membrane-envelope protein complex: structure and maturation. Science. 2008; 319(5871): 1830–4. PubMed Abstract | Publisher Full Text\n\nDokland T, Walsh M, Mackenzie JM, et al.: West Nile virus core protein; tetramer structure and ribbon formation. Structure. 2004; 12(7): 1157–63. PubMed Abstract | Publisher Full Text\n\nLuca VC, AbiMansour J, Nelson CA, et al.: Crystal structure of the Japanese encephalitis virus envelope protein. J Virol. 2012; 86(4): 2337–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArnold K, Bordoli L, Kopp J, et al.: The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics. 2006; 22(2): 195–201. PubMed Abstract | Publisher Full Text\n\nKiefer F, Arnold K, Künzli M, et al.: The SWISS-MODEL Repository and associated resources. Nucleic Acids Res. 2009; 37(Database issue): D387–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuex N, Peitsch MC, Schwede T: Automated comparative protein structure modeling with SWISS-MODEL and Swiss-PdbViewer: a historical perspective. Electrophoresis. 2009; 30(Suppl 1): S162–73. PubMed Abstract | Publisher Full Text\n\nKokernot RH, Smithburn KC, Muspratt J, et al.: Studies on arthropod-borne viruses of Tongaland. VIII. Spondweni virus, an agent previously unknown, isolated from Taeniorhynchus (Mansonioides) uniformis. S Afr J Med Sci. 1957; 22(2–3): 103–12. PubMed Abstract\n\nTognarelli J, Ulloa S, Villagra E, et al.: A report on the outbreak of Zika virus on Easter Island, South Pacific, 2014. Arch Virol. 2016; 161(3): 665–8. PubMed Abstract | Publisher Full Text\n\nBessaud M, Pastorino BA, Peyrefitte CN, et al.: Functional characterization of the NS2B/NS3 protease complex from seven viruses belonging to different groups inside the genus Flavivirus. Virus Res. 2006; 120(1–2): 79–90. PubMed Abstract | Publisher Full Text\n\nAnon. VIZIER. Reference Source\n\nBollati M, Alvarez K, Assenberg R, et al.: Structure and functionality in flavivirus NS-proteins: perspectives for drug design. Antiviral Res. 2010; 87(2): 125–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParsy CC, Alexandre FR, Bidau V, et al.: Discovery and structural diversity of the hepatitis C virus NS3/4A serine protease inhibitor series leading to clinical candidate IDX320. Bioorg Med Chem Lett. 2015; 25(22): 5427–36. PubMed Abstract | Publisher Full Text\n\nJiang Y, Andrews SW, Condroski KR, et al.: Discovery of danoprevir (ITMN-191/R7227), a highly selective and potent inhibitor of hepatitis C virus (HCV) NS3/4A protease. J Med Chem. 2014; 57(5): 1753–69. PubMed Abstract | Publisher Full Text\n\nSheng XC, Casarez A, Cai R, et al.: Discovery of GS-9256: a novel phosphinic acid derived inhibitor of the hepatitis C virus NS3/4A protease with potent clinical activity. Bioorg Med Chem Lett. 2012; 22(3): 1394–6. PubMed Abstract | Publisher Full Text\n\nSchoenfeld RC, Bourdet DL, Brameld KA, et al.: Discovery of a novel series of potent non-nucleoside inhibitors of hepatitis C virus NS5B. J Med Chem. 2013; 56(20): 8163–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKazmierski WM, Hamatake R, Duan M, et al.: Discovery of novel urea-based hepatitis C protease inhibitors with high potency against protease-inhibitor-resistant mutants. J Med Chem. 2012; 55(7): 3021–6. PubMed Abstract | Publisher Full Text\n\nRosenquist Å, Samuelsson B, Johansson PO, et al.: Discovery and development of simeprevir (TMC435), a HCV NS3/4A protease inhibitor. J Med Chem. 2014; 57(5): 1673–93. PubMed Abstract | Publisher Full Text\n\nCummings MD, Lin TI, Hu L, et al.: Discovery and early development of TMC647055, a non-nucleoside inhibitor of the hepatitis C virus NS5B polymerase. J Med Chem. 2014; 57(5): 1880–92. PubMed Abstract | Publisher Full Text\n\nLink JO, Taylor JG, Xu L, et al.: Discovery of ledipasvir (GS-5885): a potent, once-daily oral NS5A inhibitor for the treatment of hepatitis C virus infection. J Med Chem. 2014; 57(5): 2033–46. PubMed Abstract | Publisher Full Text\n\nZhang Y: I-TASSER. Reference Source\n\nZhang Y: I-TASSER server for protein 3D structure prediction. BMC Bioinformatics. 2008; 9: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang J, Yan R, Roy A, et al.: The I-TASSER Suite: protein structure and function prediction. Nat Methods. 2015; 12(1): 7–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNoble CG, Lim SP, Chen YL, et al.: Conformational Flexibility of the Dengue Virus RNA-Dependent RNA Polymerase Revealed by a Complex with an Inhibitor. J Virol. 2013; 87(9): 5291–5295. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlower DR: Bioinformatics for vaccinology. Chichester: Wiley-Blackwell, 2008. Publisher Full Text\n\nDu H, Brender JR, Zhang J, et al.: Protein structure prediction provides comparable performance to crystallographic structures in docking-based virtual screening. Methods. 2015; 71: 77–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevoin N, Calmels T, Krief S, et al.: Homology Model Versus X-ray Structure in Receptor-based Drug Design: A Retrospective Analysis with the Dopamine D3 Receptor. ACS Med Chem Lett. 2011; 2(4): 293–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarlsson J, Coleman RG, Setola V, et al.: Ligand discovery from a dopamine D3 receptor homology model and crystal structure. Nat Chem Biol. 2011; 7(11): 769–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGodbole AA, Ahmed W, Bhat RS, et al.: Targeting Mycobacterium tuberculosis topoisomerase I by small-molecule inhibitors. Antimicrob Agents Chemother. 2015; 59(3): 1549–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGodbole AA, Ahmed W, Bhat RS, et al.: Inhibition of Mycobacterium tuberculosis topoisomerase I by m-AMSA, a eukaryotic type II topoisomerase poison. Biochem Biophys Res Commun. 2014; 446(4): 916–20. PubMed Abstract | Publisher Full Text\n\nTan K, Cao N, Cheng B, et al.: Insights from the Structure of Mycobacterium tuberculosis Topoisomerase I with a Novel Protein Fold. J Mol Biol. 2016; 428(1): 182–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNguyen ED, Norn C, Frimurer TM, et al.: Assessment and challenges of ligand docking into comparative models of G-protein coupled receptors. PLoS One. 2013; 8(7): e67302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nda Silveira NJ, Arcuri HA, Bonalumi CE, et al.: Molecular models of NS3 protease variants of the Hepatitis C virus. BMC Struct Biol. 2005; 5: 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee TV, Johnson RD, Arcus VL, et al.: Prediction of the substrate for nonribosomal peptide synthetase (NRPS) adenylation domains by virtual screening. Proteins. 2015; 83(11): 2052–66. PubMed Abstract | Publisher Full Text\n\nAnon. PEPI-TiDP23-C103: First-in-Human Study to Examine the Safety, Tolerability, and Plasma Pharmacokinetics of Increasing Single and Repeated Oral Doses of TMC558445 and of a Combined Single Day Dosing of Oral TMC558445 and Oral TMC310911 and Also Oral Darunavir. 2009. Reference Source\n\nKeck ZY, Enterlein SG, Howell KA, et al.: Macaque Monoclonal Antibodies Targeting Novel Conserved Epitopes within Filovirus Glycoprotein. J Virol. 2015; 90(1): 279–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerlenghi I, Clarke M, Ruttan T, et al.: Molecular organization of a recombinant subviral particle from tick-borne encephalitis virus. Mol Cell. 2001; 7(3): 593–602. PubMed Abstract | Publisher Full Text\n\nZhang Y, Kaufmann B, Chipman PR, et al.: Structure of immature West Nile virus. J Virol. 2007; 81(11): 6141–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKostyuchenko VA, Zhang Q, Tan JL, et al.: Immature and mature dengue serotype 1 virus structures provide insight into the maturation process. J Virol. 2013; 87(13): 7700–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKostyuchenko VA, Chew PL, Ng TS, et al.: Near-atomic resolution cryo-electron microscopic structure of dengue serotype 4 virus. J Virol. 2014; 88(1): 477–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Y, Zhang W, Ogata S, et al.: Conformational changes of the flavivirus E glycoprotein. Structure. 2004; 12(9): 1607–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScheffer IE, Baraitser M, Brett EM: Severe microcephaly associated with congenital varicella infection. Dev Med Child Neurol. 1991; 33(10): 916–20. PubMed Abstract | Publisher Full Text\n\nDeasy NP, Jarosz JM, Cox TC, et al.: Congenital varicella syndrome: cranial MRI in a long-term survivor. Neuroradiology. 1999; 41(3): 205–7. PubMed Abstract | Publisher Full Text\n\nGatherer D: Zika virus protein structure homology modelling. 2016. Reference Source\n\nAnon. EMDataBank. Reference Source"
}
|
[
{
"id": "13009",
"date": "11 Apr 2016",
"name": "Andras Fiser",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper describes homology models of 15 proteins encoded in the genome of Zika virus that are built by the SWISSmodel web server. A glycolysation site was also identified. Using the models of Zika glycoprotein E a complete structural model of the virion was constructed.The practical results of the work are the 11 high quality homology models that could be used in the future for structure based drug development. I believe an interested researcher will initiate time consuming follow up studies with these models only if there is a substantial added value.These models passed quality assessment criteria according to the authors. However a number of successful models (passing similarly well the same quality requirements) can be generated that will differ in small but essential details by e.g. side chain placement, or loop conformations. However, these differences can have a dramatic effect on the outcome of subsequent drug docking trials. These alternative models can be obtained when different softwares are used, as the authors allude to it, e.g. using Modeller vs SWISSModel, as these use different forcefields and restraints to generate models. Therefore, the authors should consider providing a more insightful result by running several different modeling programs or using alternative templates (see 4., below), and comparing the results, identifying similarly modeled parts of these models and providing a set of possible solutions for subsequent studies. A battery of model quality checks were performed and additional energy minimization. However the overall high sequence identities between the target proteins and their respective templates (55% and up,) ensure that these models are highly reliable. Therefore the extensive reporting on Ramachandran plots (Figure 3) is not adding much to the results, it can be transferred to supplementary material. Similarly, Table 2 can be shrunk by eliminating many details of PROCHECK results. However the information on templates should be reported in the table, instead listing them extensively in the text. Template selection was purely based on sequence similarity requirements. Given how influential this step and how few cases require attention in this specific study, one could imagine to perform a more detailed quality check of potential templates. For instance, if several templates are available in the same sequence identity range, one should consider picking the one wit the best resolution etc. Figure 4 exposes this dilemma where two isoforms of the same template can results models with very significant differences. Energy refinement rarely, if ever, improves model quality and its use here should be better justified. Some statements require attention e.g in Abstract: “Eleven out of 15 models pass our criteria for selection”. What selection? This must be referring to quality or accuracy of models, and should be rephrased accordingly.",
"responses": [
{
"c_id": "2158",
"date": "01 Sep 2016",
"name": "Sean Ekins",
"role": "Author Response",
"response": "Dear Prof Fiser, thank you for taking the time to review and constructively comment on this manuscript. The paper describes homology models of 15 proteins encoded in the genome of Zika virus that are built by the SWISSmodel web server. A glycolysation site was also identified. Using the models of Zika glycoprotein E a complete structural model of the virion was constructed. The practical results of the work are the 11 high quality homology models that could be used in the future for structure based drug development. I believe an interested researcher will initiate time consuming follow up studies with these models only if there is a substantial added value. Response: Thank you! We would also point out the utility of this work is not only reflected in the high level of accesses, downloads and citations, but also for initiating the World Community Grid as the OpenZika project, which was launched in May, 2016 (https://www.worldcommunitygrid.org/research/zika/overview.do) These models passed quality assessment criteria according to the authors. However a number of successful models (passing similarly well the same quality requirements) can be generated that will differ in small but essential details by e.g. side chain placement, or loop conformations. However, these differences can have a dramatic effect on the outcome of subsequent drug docking trials. These alternative models can be obtained when different softwares are used, as the authors allude to it, e.g. using Modeller vs SWISSModel, as these use different force fields and restraints to generate models. Therefore, the authors should consider providing a more insightful result by running several different modeling programs or using alternative templates (see 4., below), and comparing the results, identifying similarly modeled parts of these models and providing a set of possible solutions for subsequent studies. Response: We would agree that certainly an exhaustive analysis of different methods was not undertaken. That was not our goal. We attempted to build models that we could then put into practice and provide to the community for them to use knowing the possible caveats. A battery of model quality checks were performed and additional energy minimization. However the overall high sequence identities between the target proteins and their respective templates (55% and up,) ensure that these models are highly reliable. Response: Thank you! Therefore the extensive reporting on Ramachandran plots (Figure 3) is not adding much to the results, it can be transferred to supplementary material. Response: Actually many publications use these plots and they may provide some insights on model quality so we have opted to retain them in the body of the manuscript, as the F1000Research journal does not have a limit for figures. Similarly, Table 2 can be shrunk by eliminating many details of PROCHECK results. Response: This is an internet publication we are not constrained by pages or page charges and these numbers are informative to readers. However the information on templates should be reported in the table, instead listing them extensively in the text. Response: We agree and we believe that it is neater and more accessible in a table. Therefore, we have added the information on templates in Table 2 and removed this information from the text. Template selection was purely based on sequence similarity requirements. Given how influential this step and how few cases require attention in this specific study, one could imagine to perform a more detailed quality check of potential templates. For instance, if several templates are available in the same sequence identity range, one should consider picking the one with the best resolution etc. Figure 4 exposes this dilemma where two isoforms of the same template can results models with very significant differences. Response: We totally agree – some decisions had to be made as a trade-off to get the models built in a reasonable time to put into use. We could have spent months building models by which time the x-ray structures would have been out. We had models of the proteins and the virion at least 1 month before the first cryo-EM structure. One month is a long time in drug discovery, in which we could be finding molecules for testing. Energy refinement rarely, if ever, improves model quality and its use here should be better justified. Response: We totally agree with you, Dr. Fiser. Energy minimization does not improve model quality. The word “refined”in the paper was a mistake and we have corrected it now. We are using some servers, such as KoBaMIN server, to perform structure refinement and to improve models quality, but this will be subject of another manuscript. Our main goal in this first paper was to provide the scientific community with homology models for the key proteins, which none crystal was available at the moment of the submission of this paper, and these models could be taken to initiate ZIKV antiviral drug discovery using both high throughput screens as well as structure-based design. Moreover, these structures were the core to initiate the World Community Grid project called OpenZika (https://www.worldcommunitygrid.org/research/zika/overview.do), which is a global research project to accelerate the discovery of an antiviral against the Zika virus. Some statements require attention e.g in Abstract: “Eleven out of 15 models pass our criteria for selection”. What selection? This must be referring to quality or accuracy of models, and should be rephrased accordingly. Response: OK Thank you! Changed to - Eleven out of 15 models pass our model quality criteria for their further use."
}
]
},
{
"id": "15364",
"date": "02 Aug 2016",
"name": "Bruno Botta",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe work of Sean Ekins and colleagues describes an attempt to predict the 3D structure of a number of Zika Virus proteins by homology modeling.\nInfections by Zika virus are seriously preoccupying the population and governments of endemic countries, with particular attention to pregnant women. Indeed, the most serious complications of infections by Zika virus have been observed in pregnant women (i.e. microencephalic fetus). However, the evidence that pregnant women are most affected by Zika virus infection, set some difficulties in developing focused therapeutic strategies, as nicely discussed by the authors. However, in this respect, which is the scenario proposed by authors for a candidate inhibitor of Zika virus replication. Why this molecule should be useful, and how pregnant women could deal with the administration of the drug?\nOverall, the work is well written and organized in a clear and rational way, even if no experimental validation of the work carried out in silico is proposed. Accordingly, conclusions of this work are merely speculative. My major concern is on the usability of the protein structures for drug design or virtual screening, as claimed by the authors in the abstract and discussion. Indeed, sequence identity between template and target sequences is rather low in some cases (< 60 %). Moreover, the atomistic detail of the active site and its conformation may vary noticeably in the models depending on the program used, the force field, the level of refinement (algorithm, steps, solvent model, …) and the quality of the template structure. All these variables should be taken into consideration by the authors, and the refinement of 3D models should be at least attempted or discussed in deeper details. In my personal opinion, details on the conservation rate, sequence identity and structural similarity of the binding sites would be more helpful for drug design purposes. Finally, an experimental validation of at least one of the structures modeled by authors (i.e. the most promising target for drug designing studies) should be provided. In fact, analysis with PROCHECK or the use of numerical scores is not sufficient to validate a 3D structure generated by homology modeling. There is a high risk that molecular docking towards a low-resolution structure could provide unrealistic results. The paper can be accepted after revision.",
"responses": [
{
"c_id": "2157",
"date": "01 Sep 2016",
"name": "Sean Ekins",
"role": "Author Response",
"response": "Response: Dear Prof Botta, thank you for taking the time to review and constructively comment on this manuscript. Infections by Zika virus are seriously preoccupying the population and governments of endemic countries, with particular attention to pregnant women. Indeed, the most serious complications of infections by Zika virus have been observed in pregnant women (i.e. microencephalic fetus). However, the evidence that pregnant women are most affected by Zika virus infection, set some difficulties in developing focused therapeutic strategies, as nicely discussed by the authors. However, in this respect, which is the scenario proposed by authors for a candidate inhibitor of Zika virus replication. Why this molecule should be useful, and how pregnant women could deal with the administration of the drug? Response – If we understand the question correctly - any drug for Zika should be safe for all including pregnant women. If they have a Zika infection it would be important to eradicate the virus as soon as possible so that the fetus perhaps is not exposed to the virus. Overall, the work is well written and organized in a clear and rational way, even if no experimental validation of the work carried out in silico is proposed. Response: This work was initiated in Feb 2016, months before the first crystal structures and experimental in vitro systems were published. Our goal was to build protein models that we would eventually use and which others could use in parallel, to start docking-based virtual screening and identify VS and experimental hits, to initiate the drug discovery cascade. Accordingly, conclusions of this work are merely speculative. Response: A modeling study, by definition is speculative and does not claim otherwise. However, in good faith, we provided models which did not exist. The fact that they have since gone on to form the basis of the OpenZika Project (manuscript submitted, http://openzika.ufg.br) and have been used to suggest compounds which are now being tested argues for their utility. Notwithstanding, we acknowledge that Zika protease, helicase and glycoprotein E experimental structures have now been elucidated and deposited in PDB. My major concern is on the usability of the protein structures for drug design or virtual screening, as claimed by the authors in the abstract and discussion. Indeed, sequence identity between template and target sequences is rather low in some cases (< 60 %). Response – As expected for the enzyme functions, the local active site identity and similarity scores are higher than indicated by the full-length identity %. Thus, a reasonable pocket model for docking is the primary goal, not the structural accuracy for the whole protein. We would also add that any team using our models would doubtless take a pragmatic approach that would include running various types of internal controls for virtual screening studies and confirmatory screening experiments (e.g. running the Dengue virus templates alone and checking for enrichment of established Dengue actives against the purified proteins). They would also probably prioritize their effort on the models with highest template similarity. Note also that, the first reviewer Dr. Fiser accepts that our similarity values are high enough for reliable modelling. Moreover, the atomistic detail of the active site and its conformation may vary noticeably in the models depending on the program used, the force field, the level of refinement (algorithm, steps, solvent model, …) and the quality of the template structure. Response – We agree that this is one of the challenges with this approach, but there were no other options without a single crystal structure for Zika protein when we started this work and submitted it. All these variables should be taken into consideration by the authors, and the refinement of 3D models should be at least attempted or discussed in deeper details. In my personal opinion, details on the conservation rate, sequence identity and structural similarity of the binding sites would be more helpful for drug design purposes. Response: We agree with your point. We are now refining the 3D models using molecular dynamics simulations. However, this will be the subject of work on going, taking an in depth approach to each structure and comparison to crystal structures when available. Our main goal in this paper which we submitted in Feb 2016 (and made available early March) was to provide the scientific community with the first 3D models of Zika proteins. Finally, an experimental validation of at least one of the structures modeled by authors (i.e. the most promising target for drug designing studies) should be provided. In fact, analysis with PROCHECK or the use of numerical scores is not sufficient to validate a 3D structure generated by homology modeling. Response – We agree that the experimental validation of the protein structures is of upmost importance in drug discovery programs, but this was not the scope of this paper. Again, our main goal was to generate homology models for the key proteins, which none crystal was available at the moment of the submission, and these models could be taken to initiate ZIKV antiviral drug discovery using both high throughput screens as well as structure-based design. Moreover, these structures were the core to initiate the World Community Grid project called OpenZika (https://www.worldcommunitygrid.org/research/zika/overview.do), which is a global research project to accelerate the discovery of an antiviral against the Zika virus. PROCHECK was originally developed to evaluate the stereochemical quality of experimentally determined protein structures (Laskowski R A, MacArthur M W, Moss D S, Thornton J M (1993). J. App. Cryst., 26, 283-291; Morris A L, MacArthur M W, Hutchinson E G & Thornton J M (1992). Proteins, 12, 345-364). However, PROCHECK has been widely used and reported in the literature to check the stereochemical quality of homology models, to provide numerical scores that can evaluate the statistical quality of the generated 3D protein models (See bellow the list of summarized most recent literature). It is a quality assessment criterion and is not related to experimental validation. Papers reporting the use of PROCHECK for assessing the stereochemical quality of homology models: J Biomol Struct Dyn. 2016 Apr 4:1-19 [DOI: 10.1080/07391102.2015.1117397]. Curr Cancer Drug Targets. 2015;15(9):822-35 [PMID: 26567883]. Interdiscip Sci. 2015 Aug 15. In Press [DOI: 10.1007/s12539-015-0121-z]. Onco Targets Ther. 2015 Jul 30;8:1923-30 [DOI: 10.2147/OTT.S84200]. Interdiscip Sci. 2016 Mar;8(1):41-52 [DOI: 10.1007/s12539-015-0269-6]. J Mol Model. 2015 Mar;21(3):37. [DOI: 10.1007/s00894-015-2586-4]. There is a high risk that molecular docking towards a low-resolution structure could provide unrealistic results. Response – Yes we absolutely agree, but when we started this we had no crystal structures we had to start somewhere as a means to start to filter compounds rather than random HTS. Our work has also correctly predicted the glycosylation site for glycoprotein E one of the first structures crystallized. Moreover, we have compared our models with the crystallographic structures, when they were made available. The RMSD values ranged from 0.72 to 1.8 Å, for the structures of NS1, NS3, glycoprotein E and NS2/NS3 proteins. We have also optimized the 3D models and the MolProbit scores ranged from 1.28 to 2.81 Å. MolProbit is a score that combines the clashscore, rotamer, and Ramachandran evaluations into a single score, normalized to be on the same scale as X-ray resolution. Therefore, we are optimistic that our models are reliable to start a docking-based virtual screening program in the search of a new antiviral drug, which we already initiated. The paper can be accepted after revision. Response – Thank you!"
}
]
}
] | 1
|
https://f1000research.com/articles/5-275
|
https://f1000research.com/articles/5-1610/v1
|
08 Jul 16
|
{
"type": "Research Article",
"title": "Molecular docking and molecular dynamics simulation study of inositol phosphorylceramide synthase – inhibitor complex in leishmaniasis: Insight into the structure based drug design",
"authors": [
"Vineetha Mandlik",
"Shailza Singh",
"Vineetha Mandlik"
],
"abstract": "Inositol phosphorylceramide synthase (IPCS) has emerged as an important, interesting and attractive target in the sphingolipid metabolism of Leishmania. IPCS catalyzes the conversion of ceramide to IPC which forms the most predominant sphingolipid in Leishmania. IPCS has no mammalian equivalent and also plays an important role in maintaining the infectivity and viability of the parasite. The present study explores the possibility of targeting IPCS; development of suitable inhibitors for the same would serve as a treatment strategy for the infectious disease leishmaniasis. Five coumarin derivatives were developed as inhibitors of IPCS protein. Molecular dynamics simulations of the complexes of IPCS with these inhibitors were performed which provided insights into the binding modes of the inhibitors. In vitro screening of the top three compounds has resulted in the identification of one of the compounds (compound 3) which shows little cytotoxic effects. This compound therefore represents a good starting point for further in vivo experimentation and could possibly serve as an important drug candidate for the treatment of leishmaniasis.",
"keywords": [
"IPCS (Inositol phosphorylceramide synthase)",
"Sphingolipid metabolism",
"Leishmania",
"Leishmaniasis",
"Drug Inhibitor design",
"Coumarin derivatives",
"Molecular docking",
"Molecular dynamics simulation"
],
"content": "Abbreviations\n\nIPCS – Inositol phosphorylceramide synthase, IPC – Inositol phosphorylceramide, AUR1 – Aureobasidin 1, DAG – Diacylglycerol, RMSD – Root Mean Square Deviation, LINCS – Linear constraint solver, PME – Particle Mesh Ewald.\n\n\nIntroduction\n\nInfectious disease, leishmaniasis, is the major cause of parasitic diseases affecting 12 million people worldwide. Most of the anti-leishmanial compounds do not have well-defined mechanisms. The first line treatment of cutaneous leishmaniasis involves the administration of antimony based compounds. Treatment of L. major amastigotes with Sb(V) has been found to induce apoptosis by the induction of oxidative-stress and increase in intracellular calcium1. Non-antimony based treatments such as miltefosine, and topical formulations of paromomycin are cost effective, convenient and less toxic than antimony based compounds. Amphotericin B being a liposomal formulation is expensive, has a low therapeutic index and is difficult to administer2. Newer formulations for the treatment of this disease include the administration of miltefosine. Miltefosine (hexadecylphosphocholine), originally an anticancer drug has been reported to induce apoptosis of L. major amastigotes in the infected macrophages3. Development of newer treatment modalities arise from the problem of drug resistance and quick adaptability of the parasite to the host immune response4–6.\n\nSphingolipids like IPC, form an important component of the parasitic membranes7. IPCS (inositol phosphorylceramide synthase) is an enzyme involved in the sphingolipid metabolism of protozoans and other fungal species8. The relative importance of IPCS in Leishmania has been identified through biochemical network modeling9. IPCS catalyzes the conversion of ceramide to IPC which forms the most predominant sphingolipid of the parasite10 (Figure 1). IPCS also maintains the concentration of DAG and ceramide, both of which serve as secondary messengers in several signal transduction events11. IPCS localizes into the lipid rafts of the Golgi complex12. Lipid rafts have been proposed to involve in a wide array of events like trafficking of lipid modified proteins in addition to playing an important role in the formation of signal transduction complexes13. IPCS has been important for maintaining the viability and the infectivity of several fungal species like Cryptococcus neoformans, Candida albicans and pathogens like Leishmania14–17. Interestingly there is no mammalian equivalent of this enzyme and the major sphingolipid in the host is sphingomyelin instead of IPC. Hence IPCS has been considered as a choke point enzyme in the sphingolipid metabolism of Leishmania thereby serving as a druggable target for the treatment of several fungal and protozoan diseases like leishmaniasis. LmjIPCS comprises of 338 amino acids and has 6 transmembrane domains and belongs to the PAP2c family9. IPCS is encoded by the AUR1 gene. IPCS protein present in fungi exhibits sensitivity to antifungal agents like galbonolide A, aureobasidin A, macrolidegalbonolide and khafrefungin18,19. IPCS has been recently discovered in Leishmania and to the best of our knowledge there are no reports of inhibitor design against this protein. This paper explores the possibility of targeting IPCS for the development of anti-protozoan compounds. An in silico approach for drug design has led to the development of five novel coumarin derivatives. The refinement and validation of the docked complexes has been done using molecular dynamics simulations to map the protein ligand interactions. Based on the in silico findings, the promising candidates were considered for further experimental evaluation and validation.\n\nIPCS catalyzes the reaction involving the conversion of ceramide to IPC (Inositolphosphorylceramide). IPC forms the most predominant sphingolipid in Leishmania. IPCS plays an important role in maintaining the viability of the parasite.\n\n\nMaterials and methods\n\nA set of coumarin derivatives were prepared by the assembly of pharmacophoric groups. The 2D structures of the inhibitors were drawn and edited using Chemsketch version 12.0120 (Figure 2). The SMILES format for all the compounds was generated using Open Babel version 2.3.121. Inhibitors were designed and filtered using the “Lipinski rules of five”22 and Veber’s rules23 using the Molinspiration Property Calculation Service (www.molinspiration.com).\n\nThe designed inhibitors are Coumarin derivatives. Coumarin increases the phagocytic activity of the macrophages.\n\nThe pharmacophore models describing the inherent chemical features of the inhibitors were generated using the “Feature mapping protocol” available in Discovery Studio version 3.0. (www.accelyrs.com). Pharmacophore models of the inhibitors indicated that the ligand had at least a maximum of 5 pharmacophoric features i.e. Hydrogen bond acceptors (HBA), Hydrogen bond donors (HBD), positive ionizable groups (PI), Ring aromatic groups (RA) and the Hydrophobic groups (HY) present in the ligand.\n\nIPCS is one of the emerging drug targets for the treatment of leishmaniasis. The crystal structure of the IPCS protein has not been solved and hence the 3D structure for the IPCS protein developed by our group before has been used for the inhibitor design. The model was developed using the I-TASSER server (http://zhanglab.ccmb.med.umich.edu/I-TASSER/). The predicted model has a total of 338 amino acid residues and has 7 transmembrane helices9. The 3D structure of IPCS was energy minimized by the steepest gradient method of energy minimization using the GROMACS 4.0 package24. Mol2 file format of the inhibitors was converted to PDBQT format using MGL tools prior to docking. All the water and solvent atoms of the protein were removed prior to docking and the polar hydrogens were added. The protein was kept rigid while the ligand was allowed to rotate and explore more flexible binding pockets. Docking of the inhibitors onto the IPCS protein was performed using Auto Dock 4 version 1.5.6 and Auto Dock vina. version 1.1.2. The grid box size dimensions were 40X40X40, the default scoring function was used for docking25,26. Binding modes of the docked complexes were obtained and the amino acid residues present at a distance of 5Å were considered as the binding partners of the ligands. The interaction diagrams representing the docked complexes have been generated using Pymol v 1.3.\n\nMolecular dynamics simulation is a computational method that provides information regarding the time dependent behavior of any molecular system by integrating Newton’s laws of motion. The docked complexes (IPCS-inhibitor complex) were subjected to MD simulation using Desmond version 4.4 (Schodinger Biosuite). MD simulation of both the IPCS protein and IPCS –ligand complexes were performed for a time period of 10ns by using the OPLS force field. The complex was centered in a cubic box and filled with TIP3P water molecules. The system was neutralized and the initial energy minimization for the system was done using the conjugant gradient algorithm. The Martyna-Tobias-Klein scheme was used for pressure coupling. Electrostatic forces were calculated using the PME algorithm27. All runs were performed at 300K at constant volume and temperature (NPT ensemble) under certain periodic boundary conditions. RMSD plots for the backbone atoms for both the protein and ligand bound protein were generated to understand the relative stability of the ligand inside its binding pocket and the IPCS-inhibitor complexes were visualized.\n\nMacrophage cell population was collected post 24 h treatment with the compound 3, washed and suspended in 1XPBS. Cells were stained with 10µl of 10μg/mL of propidium iodide (PI) dye (Invitrogen) and acquired on FACS. Total macrophage population was gated based on their forward scatter (FSC) and side scatter (SSC) after excluding the cell debris. A minimum of 10,000 events were acquired for each sample on FACS Canto II (Beckon Dickson, San Jose, California) and analyzed using FACS Diva Software (version 6.2.1) (Beckon Dikson, San Jose, California).\n\n\nResults\n\nA group of coumarin derivatives were prepared as inhibitors of the IPCS protein belonging to L. major. Assessment of the drug like properties indicated that all the inhibitors were found to comply with the Lipinski’s “Rule of five” (molecular weight (Mwt) ≤ 500, clogP ≤ 5, H-bond donors (HBD) ≤ 5, and acceptors (HBA) ≤ 10) and Verber’s rules (no. of rotatable bonds < 10, PSA ≤ 140A2) (Table 1).\n\nHBA – Hydrogen bond acceptor, HBD – Hydrogen bond donor, HY – Hydrophobic, RA – Ring aromatic, MR – Molar refractivity, NROTB – No. of rotatable bonds, cLogP – log octanol/water partition coefficient, PSA – Polar surface area, NSC – No. of stereo centers.\n\nMolecular docking studies reveal the binding modes of the ligand with the IPCS protein giving an insight into the crucial amino acid residues that are involved during the binding. A comparison of the binding energies of all the compounds indicates that compound 3 has the least binding energy among all and hence exhibits maximum affinity towards the IPCS protein (Table 2). The interaction modes of all the IPCS inhibitors post docking along with their pharmacophoric features have been presented [Figure 3]. Binding mode analysis reveals that hydrophilic amino acids like Arg299 and His220 were found to be involved in hydrogen or π bonding with most of the ligands (Table 3). The relative stability of the compounds within the binding site was maintained due to the van der Waal’s interaction between the hydrophobic amino acids of the IPCS protein and the ligand (Table 4).\n\nThe pharmacophoric features such as hydrogen bond acceptors (green), hydrogen bond donors (pink), hydrophobic regions (blue) and the aromatic rings in yellow are shown in the figure.\n\nProtein backbone RMSD plots indicate the stability of the IPCS-inhibitor complex. The drug backbone RMSD plots indicate that compounds 2 and 3 maintained their interactions with the IPCS protein (Figure 4). Binding modes of compounds 1 to 5 post MD simulation have been shown in Figure 5a–e.\n\nBackbone RMSD of a) Compound 1 and b) Compound 2 c) Compound 3 d) Compound 4 e) Compound 5 is shown in the figure. Compound 1, 2 and 3 appear to maintain their stability within the binding pocket as they show lower RMSD fluctuations.\n\nThe interaction of the ligand within the IPCS inhibitor complex post MD simulation is shown the figure a) IPCS - compound 1 complex b) IPCS - compound 2 complex c) IPCS - compound 3 complex d) IPCS - compound 4 complex and e) IPCS - compound 5 complex. MD simulation was performed for a time period of 10ns. Interacting residues are represented in red.\n\nThe cytotoxicity profile of compound 3 was checked over the macrophage cell line. Of all five compounds, compound 3 had the highest viability. The viability of C3 treated macrophages (67.3%) was slightly lesser than the control (73.5%).\n\n\nDiscussion\n\nIPCS (Inositol phosphorylceramide synthase) has been identified as an important drug target in the sphingolipid metabolism of several organisms like fungi, yeast and protozoans like Leishmania and Trypanosoma28. Systems biology has played a major role in defining the relative importance of IPCS in the sphingolipid metabolism of Leishmania, a protozoan responsible for causing an infectious disease leishmaniasis. The quest for developing new inhibitors for any target protein relies mainly on in silico approaches like computer based docking which involves the generation of a comprehensive set of ligand conformations that are eventually scored and ranked according to their stability and affinity for the protein. Coumarin has been shown to simulate the macrophages, enhancing their phagocytic ability29. A total of five ligands were developed as inhibitors for the IPCS protein. Molecular docking of the inhibitors with the IPCS protein revealed the binding modes of inhibitors. To account for the flexibility of the protein and ligand and to determine the binding affinity of the inhibitors with the IPCS protein, a 10 ns molecular dynamics simulation of the docked complexes was carried out. Binding mode analysis revealed that the binding modes obtained after MD simulation were more or less similar to that obtained post docking (Table 4). The presence of a large number of H bond acceptors, H bond donors as well as hydrophobic groups in the ligands account for the stability of the ligand inside the binding pocket of IPCS. Based on the RMSD of the ligand-protein complex, it was observed that compounds 1, 2 and 3 maintained their interaction with the protein with lower RMSD fluctuations. Out of these, compound 3 showed the highest binding affinity and its cytotoxicity was assessed using flow cytometry. Cytotoxicity of compound 3 was lesser as compared to other compound. A comparison of compound 3 treated macrophages along with the untreated macrophages has been made in Figure 6.\n\nMacrophages were treated with compound 3 for 24h. a) Control cell population displayed a percentage viability of 73.5% b) Compound 3 (1mg/ml) treated macrophages displayed a viability of 67.3% post 24hr treatment.\n\n\nConclusion\n\nThere is an urgent need to design and develop novel anti-leishmanial compounds due to various problems associated with the current chemotherapeutics for the treatment of this disease. IPCS has been proposed to be a probable drug target in the sphingolipid pathway of Leishmania. We have designed a few novel coumarin derivatives using in silico approaches. MD simulation post docking studies reveal the interactions between the IPCS protein and ligands. Binding modes obtained after docking and after MD simulation reveal almost identical binding modes which is suggestive of the selectivity and selectivity of the ligand towards the active site of the IPCS protein.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for ‘Molecular docking and molecular dynamics simulation study of inositol phosphorylceramide synthase – inhibitor complex in leishmaniasis: Insight into the structure based drug design’, 10.5256/f1000research.9151.d12833730",
"appendix": "Author contributions\n\n\n\nSS designed and conceptualized the experiments. VM carried out the experiments. SS and VM wrote the manuscript. All the authors have read and approved the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe present work is being funded by the Department of Biotechnology (DBT) Project No: BT/PR 6037/GBD/27/372/2012.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nThe authors would like to thank the Director, National Center for Cell Science (NCCS) for supporting the Bioinformatics and High Performance Computing Facility (BHPCF) at NCCS, Pune, India.\n\n\nReferences\n\nHaldar AK, Sen P, Roy S: Use of antimony in the treatment of leishmaniasis: current status and future directions. Mol Biol Int. 2011; 2011: 23, 571242. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYardley V, Croft SL: Activity of liposomal amphotericin B against experimental cutaneous leishmaniasis. Antimicrob Agents Chemother. 1997; 41(4): 752–756. PubMed Abstract | Free Full Text\n\nKhademvatan S, Gharavi MJ, Rahim F, et al.: Miltefosine-Induced Apoptotic Cell Death on Leishmania major and L. tropica Strains. Korean J Parasitol. 2011; 49(1): 17–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHadighi R, Mohebali M, Boucher P, et al.: Unresponsiveness to Glucantime treatment in Iranian cutaneous leishmaniasis due to drug-resistant Leishmania tropica parasites. PLoS Med. 2006; 3(5): e162. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMishra BB, Kale RR, Singh RK, et al.: Alkaloids: future prospective to combat leishmaniasis. Fitoterapia. 2009; 80(2): 81–90. PubMed Abstract | Publisher Full Text\n\nCroft SL, Barrett MP, Urbina JA: Chemotherapy of trypanosomiases and leishmaniasis. Trends Parasitol. 2005; 21(11): 508–512. PubMed Abstract | Publisher Full Text\n\nHeung LJ, Luberto C, Del Poeta M: Role of sphingolipids in microbial pathogenesis. Infect Immun. 2006; 74(1): 28–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLester RL, Dickson RC: Sphingolipids with inositolphosphate-containing head groups. Adv Lipid Res. 1993; 26: 253–274. PubMed Abstract\n\nMandlik V, Shinde S, Chaudhary A, et al.: Biological network modeling identifies IPCS in Leishmania as a therapeutic target. Integr Biol (Camb). 2012; 4(9): 1130–1142. PubMed Abstract | Publisher Full Text\n\nDickson RC: Sphingolipid functions in Saccharomyces cerevisiae: comparison to mammals. Annu Rev Biochem. 1998; 67: 27–48. PubMed Abstract | Publisher Full Text\n\nCerbón J, Falcon A, Hernández-Luna C, et al.: Inositol phosphoceramide synthase is a regulator of intracellular levels of diacylglycerol and ceramide during the G1 to S transition in Saccharomyces cerevisiae. Biochem J. 2005; 388(Pt 1): 169–176. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDenny PW, Shams-Eldin H, Price HP, et al.: The protozoan inositol phosphorylceramide synthase: a novel drug target that defines a new class of sphingolipid synthase. J Biol Chem. 2006; 281(38): 28200–28209. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimons K, Sampaio JL: Membrane organization and lipid rafts. Cold Spring Harb Perspect Biol. 2011; 3(10): a004697. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenry J, Guillotte A, Luberto C, et al.: Characterization of inositol phospho-sphingolipid-phospholipase C 1 (Isc1) in Cryptococcus neoformans reveals unique biochemical features. FEBS Lett. 2011; 585(4): 635–640. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWells GB, Dickson RC, Lester RL: Isolation and composition of inositolphosphorylceramide-type sphingolipids of hyphal forms of Candida albicans. J Bacteriol. 1996; 178(21): 6223–6226. PubMed Abstract | Free Full Text\n\nLevine TP, Wiggins CA, Munro S: Inositol phosphorylceramide synthase is located in the Golgi apparatus of Saccharomyces cerevisiae. Mol Biol Cell. 2000; 11(7): 2267–2281. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGüther ML, Lee S, Tetley L, et al.: GPI-anchored proteins and free GPI glycolipids of procyclic form Trypanosoma brucei are nonessential for growth, are required for colonization of the tsetse fly, and are not the only components of the surface coat. Mol Biol Cell. 2006; 17(12): 5265–5274. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNagiec MM, Nagiec EE, Baltisberger JA, et al.: Sphingolipid synthesis as a target for antifungal drugs. Complementation of the inositol phosphorylceramide synthase defect in a mutant strain of Saccharomyces cerevisiae by the AUR1 gene. J Biol Chem. 1997; 272(15): 9809–9817. PubMed Abstract | Publisher Full Text\n\nYano T, Aoyagi A, Kozuma S, et al.: Pleofungins, novel inositol phosphorylceramide synthase inhibitors, from Phoma sp. SANK 13899. I. Taxonomy, fermentation, isolation, and biological activities. J Antibiot (Tokyo). 2007; 60: 136–142. PubMed Abstract | Publisher Full Text\n\nACD/Chemsketch, version 5.12. Advanced Chemistry Development, Inc., Toronto, ON, Canada. Reference Source\n\nO'Boyle NM, Banck M, James CA, et al.: Open Babel: An open chemical toolbox. J Cheminform. 2011; 3: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLipinski CA, Lombardo F, Dominy BW, et al.: Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 1997; 23(1-3): 3–25. PubMed Abstract | Publisher Full Text\n\nVeber DF, Johnson SR, Cheng HY, et al.: Molecular properties that Influence the oral bioavailability of drug candidates. J Med Chem. 2002; 45(12): 2615–2623. PubMed Abstract | Publisher Full Text\n\nBerendsen HJ, van der Spoel D, van Drunen R: GROMACS: a message-passing parallel molecular dynamics implementation. Comp Phys Comm. 1995; 91(1–3): 43–56. Publisher Full Text\n\nMorris GM, Goodsell DS, Halliday RS, et al.: Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Computational Chemistry. 1998; 19(14): 1639–1662. Publisher Full Text\n\nTrott O, Olson AJ: AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem. 2010; 31(2): 455–461. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDarden T, York D, Pedersen L: Particle mesh Ewald: an N-log(N) method for Ewald sums in large systems. J Chem Phys. 1993; 98(12): 10089–10092. Publisher Full Text\n\nZhang K, Beverley SM: Phospholipid and sphingolipid metabolism in Leishmania. Mol Biochem Parasitol. 2010; 170(2): 55–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPiller NB: A morphological assessment of the stimulatory effect of coumarin on macrophages. Br J Exp Pathol. 1978; 59(1): 93–96. PubMed Abstract | Free Full Text\n\nMandlik V, Singh S: Dataset 1 in: Molecular docking and molecular dynamics simulation study of inositol phosphorylceramide synthase – inhibitor complex in leishmaniasis: Insight into the structure based drug design. F1000Research. 2016. Data Source"
}
|
[
{
"id": "14858",
"date": "20 Jul 2016",
"name": "François Ferron",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMandlik and Singh are presenting a very interesting study against Leishmania. Using in silico methods they have identified promising compounds.\nThey chose with great reason IPCS as a target for its uniqueness to the pathogens the study very interesting yet the manuscript needs some clarifications.\nAs there is no structure the authors have done a model, the model should be presented here and I feel a previous reference from 2012 won't do. The reader needs to be introduced to it and at least to have a clear understanding of the catalytic site and docking site.\n\nIn the material and methods it would be appreciated to have an idea where is the center of the grid (may be with a figure). Can you explain why you chose a grid with large dimensions?\n\nThe analysis of the viability raises a question. Why the surface p1 is not the same between control and compound? To compare the stats it should be the same size here it is half. Also from fig 6 compound 3 it seems that a lot of data were excluded from P1 area? Can you explain?\n\nAs perspective are you planning on testing in vitro the efficiency of compound 3 and have an idea of binding affinity?\nMinor comment on figure 2: structures of compounds are distorted and could you put all the compound in same orientation 1 (and 2 are upside down). It will help to compare the geometry and differences between molecules.\nFigure 5: description of interactions could be better represented ma be with LigPlot, as it is it is not clear.",
"responses": [
{
"c_id": "2120",
"date": "01 Sep 2016",
"name": "Shailza Singh",
"role": "Author Response",
"response": "As there is no structure the authors have done a model, the model should be presented here and I feel a previous reference from 2012 won't do. The reader needs to be introduced to it and at least to have a clear understanding of the catalytic site and docking site. Author’s response: The authors have accepted the suggestion made and have now included a figure showing the binding cavity that was predicted for the IPCS protein around which the grid box was centered during docking. In the material and methods it would be appreciated to have an idea where the center of the grid is (may be with a figure). Author’s response: Figure 3 has been included keeping in mind the suggestion made. Can you explain why you chose a grid with large dimensions? Author’s response: The inhibitors designed have not yet been reported. As there are no studies indicating the exact binding site in the IPCS protein, we have made binding site prediction and the grid box dimensions have been adjusted to incorporate most of the amino acids that fall in the binding site. The analysis of the viability raises a question. Why the surface p1 is not the same between control and compound? To compare the stats it should be the same size here it is half. Also from fig 6 compound 3 it seems that a lot of data were excluded from P1 area? Can you explain? Author’s response: The P1 area has been demarcated as per the untreated macrophages (control). As was a decrease in granularity of macrophages post treatment, the cells had lower SSC, however the viability of the cells has not decreased. As per the author’s knowledge about flow cytometry, the P1 area has to remain the same both for the control and treated samples. As perspective are you planning on testing in vitro the efficiency of compound 3 and have an idea of binding affinity? Author’s response: At present, the authors don’t have idea of the binding affinity of compound 3. Such studies can be done in future. Minor comment on figure 2: structures of compounds are distorted and could you put all the compound in same orientation 1 (and 2 are upside down). It will help to compare the geometry and differences between molecules. Author’s response: Changes suggested have been made in the Figure 2. Figure 5: description of interactions could be better represented maybe with LigPlot, as it is it is not clear. Author’s response: Changes suggested have been made in the Figure 6. We thank the reviewer for his valuable suggestions which went a long way in improving the said manuscript."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1610
|
https://f1000research.com/articles/5-2121/v1
|
31 Aug 16
|
{
"type": "Opinion Article",
"title": "Plant expression systems, a budding way to confront chikungunya and Zika in developing countries?",
"authors": [
"Jaime A. Cardona-Ospina",
"Juan C. Sepúlveda-Arias",
"L. Mancilla",
"Luis G. Gutierrez-López",
"Juan C. Sepúlveda-Arias",
"L. Mancilla",
"Luis G. Gutierrez-López"
],
"abstract": "Plant expression systems could be used as biofactories of heterologous proteins that have the potential to be used with biopharmaceutical aims and vaccine design. This technology is scalable, safe and cost-effective and it has been previously proposed as an option for vaccine and protein pharmaceutical development in developing countries. Here we present a proposal of how plant expression systems could be used to address Zika and chikungunya outbreaks through development of vaccines and rapid diagnostic kits.",
"keywords": [
"Chikungunya",
"Zika",
"Molecular farming",
"Vaccine"
],
"content": "\n\nPlant expression systems have been used for the past 26 years for the production of human or animal proteins of biopharmaceutical interest. Antigens, antibodies, and enzymes have been produced, and some of them commercialized using several plant expression platforms1. While certain impediments remain such as community reluctance to accept transgenic products and the strict regulations for approval2,3, the FDA approval of ELELYSO® (alfa taliglucerase), a recombinant cerebrosidase for the treatment of Gaucher’s disease, has motivated plant-based biopharmaceutical protein production. These methods are all the more attractive because they are cost-effective, safe and scalable4.\n\nAfter the arrival of Zika and chikungunya viruses to Latin American countries, they quickly became endemic diseases. They currently pose an acute and chronic burden for health systems and represent a diagnostic challenge in areas where those infections co-circulate with dengue and other febrile-illnesses5,6. Clinical diagnosis frequently is difficult given the similar clinical features with other viral infections such as dengue. However, the laboratory confirmation of Zika, dengue or chikungunya infection is important because each one has different implications for follow-up both in the short and long term. Diagnosis of acute, symptomatic infection is typically achieved through pathogen detection by virus isolation or qRT-PCR, Serology may be helpful later in the acute illness, but requires convalescent sampling in many cases and comes at a significant cost for healthcare systems. For this reason, confirmation is not recommended for the general population and has been restricted to specific cases. On the other hand, prevention of infection has been in the spotlight for policy makers. There are Zika and chikungunya vaccines under development, but current vaccine production is compromised by reduced capacity of vaccine manufacturers and substantial unmet needs for investment7.\n\nDeveloping countries have been the most affected worldwide with these vector-borne diseases, and plant-based expression platforms have been proposed as a biotechnological tool to address the vaccine development challenge8. Plants could be used as bio-factories for the production of antigens, for both rapid diagnostic test design and vaccine production. Plant platforms operate at a small fraction of the cost (0.1% to 10%) of other expression systems like bacteria or mammalian cells9. Additionally, it has higher protein yield, lower contamination risk, lower storage cost, ability to assemble complex proteins with minor glycosylation differences, as well as high product quality, safety, and scalability9,10.\n\nAlthough post-translational modifications have been a concerning issue, plant-derived vaccines can elicit protective immune responses11,12. Glycoengineering allows modification of protein glycosylation patterns in order to improve immunogenicity. Additionally, plant derived polysaccharides have been proposed as adjuvants and vehicles, further highlighting plants as a biofactory for antigen production10,11. Finally, viral antigens produced in plants have been used to target other arboviruses like the West Nile virus11.\n\nIn this context, a research agenda to assess the production of pharmaceutical proteins through plant molecular farming seems like a possible scenario to deal with current arboviruses epidemics. At first, candidate proteins should be defined. These proteins should be highly conserved and highly immunogenic. Importantly, antigen similarity between flaviviruses like dengue, yellow fever and Zika viruses has limited target antigen selection, for both vaccine and diagnostic test design for Zika, because of cross-reactivity and the risk of antibody dependent enhancement of infection13. Regarding chikungunya virus, the envelope glycoprotein E2 has been studied for both vaccine and rapid diagnostic test design, and even the use of plant produced virus like particles has been proposed as candidate for vaccine production14 (Table 1).\n\n*Plant glycosylation could both enhance or limit immune response.\n\nIn addition, the plant expression system should be carefully selected. It is important to note that not all plants can be transformed, and phenolic compounds produced by plants and some of its secondary metabolites could make the purification of the desired protein difficult. Furthermore, the risk of contamination of other crops by the spread of transgenic pollen must be monitored according to the transformation method and plant species used for it. Transformation protocols in Nicotiana benthamiana, N. tabacum, and Solanum tuberosum have been used most commonly9.\n\nBecause of its scalability, efficiency and effectiveness, transformation using A. tumefaciens has been the preferred method for biopharmaceutical protein production. This method does not require special equipment like the gene gun, it allows a more precise and selective transgene insertion, and results in lower tissue damage and thus higher available biomass for protein production. Using this method both transient and stable transformation is obtained. In recent years, the method of agroinfiltration for transient plant transformation is preferred because of its potential to be systematized and provide an adequate yield of protein in the short-term4,9.\n\nIn conclusion, plant expression systems of heterologous proteins are a feasible strategy for vaccine development and rapid diagnostic kit design. Additionally, it could enable developing countries to address the challenge of current arboviruses epidemics, both in improving diagnostics as well as increasing primary prevention. The development of a molecular plant farming research agenda seems as a worthy solution to empower research in developing countries. It will permit every country to take advantage of its own natural resources in an individualized manner to deal with its own epidemiologic challenges.",
"appendix": "Author contributions\n\n\n\nJACO and JCSA conceived the idea. JACO, JCSA, LM and LGGL carried out the literature search. JACO and JCSA prepared the first draft. LM and LGGL contributed to its revision. All authors were involved in the revision of the final draft of the manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was funded by Sistema General de Regalías de Colombia and Universidad Tecnológica de Pereira, assigned to Jaime A. Cardona-Ospina in the framework of the project \"Development of scientific and technological skills in biotechnology applied to health and agro-industry sectors in Risaralda\" (BPIN 2012000100050).\n\n\nAcknowledgements\n\nWe thank Dr. Matthew Collins (Division of Infectious Diseases, University of North Carolina) for assisting us with language style.\n\n\nReferences\n\nLau OS, Sun SS: Plant seeds as bioreactors for recombinant protein production. Biotechnol Adv. 2009; 27(6): 1015–22. PubMed Abstract | Publisher Full Text\n\nMor TS: Molecular pharming's foot in the FDA's door: Protalix's trailblazing story. Biotechnol Lett. 2015; 37(11): 2147–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpök A, Twyman RM, Fischer R, et al.: Evolution of a regulatory framework for pharmaceuticals derived from genetically modified plants. Trends Biotechnol. 2008; 26(9): 506–17. PubMed Abstract | Publisher Full Text\n\nChen Q, Lai H, Hurtado J, et al.: Agroinfiltration as an Effective and Scalable Strategy of Gene Delivery for Production of Pharmaceutical Proteins. Adv Tech Biol Med. 2013; 1(1): pii: 103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCardona-Ospina JA, Villamil-Gómez WE, Jimenez-Canizales CE, et al.: Estimating the burden of disease and the economic cost attributable to chikungunya, Colombia, 2014. Trans R Soc Trop Med Hyg. 2015; 109(12): 793–802. PubMed Abstract | Publisher Full Text\n\nVillamil-Gómez WE, González-Camargo O, Rodriguez-Ayubi J, et al.: Dengue, chikungunya and Zika co-infection in a patient from Colombia. J Infect Public Health. 2016; 9(5): 684–6. PubMed Abstract | Publisher Full Text\n\nPoland GA, Whitaker JA, Poland CM, et al.: Vaccinology in the third millennium: scientific and social challenges. Curr Opin Virol. 2016; 17: 116–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHefferon K: Plant-derived pharmaceuticals for the developing world. Biotechnol J. 2013; 8(10): 1193–202. PubMed Abstract | Publisher Full Text\n\nYao J, Weng Y, Dickey A, et al.: Plants as Factories for Human Pharmaceuticals: Applications and Challenges. Int J Mol Sci. 2015; 16(12): 28549–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosales-Mendoza S, Salazar-González JA, Decker EL, et al.: Implications of plant glycans in the development of innovative vaccines. Expert Rev Vaccines. 2016; 15(7): 915–25. PubMed Abstract | Publisher Full Text\n\nHe J, Peng L, Lai H, et al.: A plant-produced antigen elicits potent immune responses against West Nile virus in mice. Biomed Res Int. 2014; 2014: 952865. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPineo CB, Hitzeroth II, Rybicki EP: Immunogenic assessment of plant-produced human papillomavirus type 16 L1/L2 chimaeras. Plant Biotechnol J. 2013; 11(8): 964–75. PubMed Abstract | Publisher Full Text\n\nMartins KA, Dye JM, Bavari S: Considerations for the development of Zika virus vaccines. Vaccine. 2016; 34(33): 3711–2. PubMed Abstract | Publisher Full Text\n\nSalazar-Gonzalez JA, Angulo C, Rosales-Mendoza S: Chikungunya virus vaccines: Current strategies and prospects for developing plant-made vaccines. Vaccine. 2015; 33(31): 3650–8. PubMed Abstract | Publisher Full Text\n\nKhan M, Dhanwani R, Rao PV, et al.: Subunit vaccine formulations based on recombinant envelope proteins of Chikungunya virus elicit balanced Th1/Th2 response and virus-neutralizing antibodies in mice. Virus Res. 2012; 167(2): 236–46. PubMed Abstract | Publisher Full Text\n\nVerma A, Chandele A, Nayak K, et al.: High yield expression and purification of Chikungunya virus E2 recombinant protein and its evaluation for serodiagnosis. J Virol Methods. 2016; 235: 73–9. PubMed Abstract | Publisher Full Text\n\nWeber C, Büchner SM, Schnierle BS: A small antigenic determinant of the Chikungunya virus E2 protein is sufficient to induce neutralizing antibodies which are partially protective in mice. PLoS Negl Trop Dis. 2015; 9(4): e0003684. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBao H, Ramanathan AA, Kawalakar O, et al.: Nonstructural protein 2 (nsP2) of Chikungunya virus (CHIKV) enhances protective immunity mediated by a CHIKV envelope protein expressing DNA Vaccine. Viral Immunol. 2013; 26(1): 75–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChattopadhyay A, Wang E, Seymour R, et al.: A chimeric vesiculo/alphavirus is an effective alphavirus vaccine. J Virol. 2013; 87(1): 395–402. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChang LJ, Dowd KA, Mendoza FH, et al.: Safety and tolerability of chikungunya virus-like particle vaccine in healthy adults: a phase 1 dose-escalation trial. Lancet. 2014; 384(9959): 2046–52. PubMed Abstract | Publisher Full Text\n\nAkahata W, Yang ZY, Andersen H, et al.: A virus-like particle vaccine for epidemic Chikungunya virus protects nonhuman primates against infection. Nat Med. 2010; 16(3): 334–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHallengard D, Lum FM, Kummerer BM, et al.: Prime-boost immunization strategies against Chikungunya virus. J Virol. 2014; 88(22): 13333–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFric J, Bertin-Maghit S, Wang CI, et al.: Use of human monoclonal antibodies to treat Chikungunya virus infection. J Infect Dis. 2013; 207(2): 319–22. PubMed Abstract | Publisher Full Text\n\nGoh LY, Hobson-Peters J, Prow NA, et al.: Monoclonal antibodies specific for the capsid protein of chikungunya virus suitable for multiple applications. J Gen Virol. 2015; 96(Pt 3): 507–12. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "16476",
"date": "28 Sep 2016",
"name": "Antonio Carlos Albuquerque Bandeira",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDr Ospina gives us a clear insight on the plant platforms for producing antigens for both diagnosis and for vaccine development in Chikungunya and Zika. It enables a large scale production with lower costs and should be evaluated in countries with resource-limiting conditions.",
"responses": []
},
{
"id": "16839",
"date": "06 Oct 2016",
"name": "Fredrik Pettersson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors discuss an interesting concept of using plants as vehicles for the production of vaccines and pharmaceutically important proteins. Although it is a promising field with some successful examples already existing, I get the impression from the text that it may lead to a quick solution for the vaccine development targeting the current Zika and chikungunya outbreaks. However, given the time for the development and testing needed before such a vaccine could be administered to a larger population I cannot foresee that this will happen in the near future. Maybe this can be rephrased to reflect this. Furthermore, the skeptical or negative opinion against transgenic crops existing in many countries may also affect the medical substances in question here, especially if targeting edible vaccines. As this may be a larger obstacle than the actual technological challenges, this could have been discussed further.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2121
|
https://f1000research.com/articles/5-2107/v1
|
30 Aug 16
|
{
"type": "Research Article",
"title": "Changes in labial capillary density on ascent to and descent from high altitude",
"authors": [
"Edward Gilbert-Kawai",
"Jonny Coppel",
"Phillip Hennis",
"Michael P. Grocott",
"Can Ince",
"Daniel Martin",
"Jonny Coppel",
"Phillip Hennis",
"Michael P. Grocott",
"Can Ince",
"Daniel Martin"
],
"abstract": "Present knowledge of how the microcirculation is altered by prolonged exposure to hypoxia at high altitude is incomplete and modification of existing analytical techniques may improve our knowledge considerably. We set out to use a novel simplified method of measuring in vivo capillary density during an expedition to high altitude using a CytoCam incident dark field imaging video-microscope. The simplified method of data capture involved recording one-second images of the mucosal surface of the inner lip to reveal data about microvasculature density in ten individuals. This was done on ascent to, and descent from, high altitude. Analysis was conducted offline by two independent investigators blinded to the participant identity, testing conditions and the imaging site. Additionally we monitored haemoglobin concentration and haematocrit data to see if we could support or refute mechanisms of altered density relating to vessel recruitment. Repeated sets of paired values were compared using Kruskall Wallis Analysis of Variance tests, whilst comparisons of values between sites was by related samples Wilcoxon Signed Rank Test. Correlation between different variables was performed using Spearman’s rank correlation coefficient, and concordance between analysing investigators using intra-class correlation coefficient. There was a significant increase in capillary density from London on ascent to high altitude; median capillaries per field of view area increased from 22.8 to 25.3 (p=0.021). There was a further increase in vessel density during the six weeks spent at altitude (25.3 to 32.5, p=0.017). Moreover, vessel density remained high on descent to Kathmandu (31.0 capillaries per field of view area), despite a significant decrease in haemoglobin concentration and haematocrit. Using a simplified technique, we have demonstrated an increase in capillary density on early and sustained exposure to hypobaric hypoxia at thigh altitude, and that this remains elevated on descent to normoxia. The technique is simple, reliable and reproducible.",
"keywords": [
"Capillaries",
"Microcirculation",
"Altitude",
"Microscopy",
"Oxygen"
],
"content": "List of Abbreviations\n\nEBC Everest Base Camp\n\nFOV Field of view area\n\n[Hb] Haemoglobin concentration\n\nHct Haematocrit\n\nIDF Incident Dark Field\n\nKTM Kathmandu\n\nLON London\n\nSpO2 Peripheral oxygen saturation\n\n\nIntroduction\n\nThe physiological processes involved in acclimatisation to high altitude attempt to maintain adequate oxygen delivery as the partial pressure of oxygen decreases. Traditionally, research has concentrated on global haemodynamics and the macrocirculation, variables such as cardiac output1, oxygen saturations2 and haemoglobin concentration [Hb]3. Far fewer studies have focused on the microcirculation, which regulates blood flow to match micro-regional oxygen demand. Disruption of microvascular blood flow could explain a failure of acclimatisation in some individuals as well as the well-documented exercise limitation that occurs at altitude despite normalisation of systemic oxygen delivery4. The precise role of the microcirculation in acclimatisation to hypoxia, however, remains unclear.\n\nTeleological reasoning would suggest that increasing capillary density could provide a means to augment oxygen flux and tissue oxygenation through a reduction in the inter-capillary distance5. Whilst plausible, data on this theory remains contradictory, though this may in part relate to the dissimilar tissues observed. In human skeletal muscle biopsy samples previously exposed to hypobaric hypoxia, no evidence of neovascularisation has been demonstrated6–9. Interestingly, in each instance whereby the capillary density was initially thought to increase, no change in the capillary-to-fibre ratio was observed. The perceived rise in capillary density were therefore interpreted as being secondary occurrences in response to a reduction in skeletal muscle mass. Conversely, an increase in the density of sublingual microcirculatory vessels to >25 μm was demonstrated on ascent to high altitude10,11, a response that was further amplified after prolonged exposure to hypoxia10. In this instance, what remains to be determined is whether the observed changes in vessel density are due to microvascular recruitment secondary to increased blood viscosity (and thus quickly reversible), or neovascularization (which is likely to be sustained). Moreover, the question of what happens to vessel density following re-exposure to normoxia remains to be elucidated.\n\nWe therefore piloted a novel modification of a previously described technique for calculating changes in capillary density12,13 on ten individuals, to see if we could firstly support or refute previous findings on ascent to high altitude, and secondly see if the changes observed persist on descent. Additionally we monitored haemoglobin concentration and haematocrit data to see if we could support of refute mechanisms of altered density relating to vessel recruitment.\n\n\nMethods\n\nThe study was undertaken as part of the Xtreme Everest 2 research expedition (XE2)14. The study design, risk management plan and protocol were approved (in accordance with the declaration of Helsinki) both by the University College London Committee and the Ethics of Non-National Health Service Human Research, and the Nepal Health Research Council (Reg no. 139/2012). Written consent was obtained from all participants. Baseline images of the labial capillaries were initially obtained from ten individuals in London (LON) (35m) in December 2012 and January 2013. Sequential images were taken after an 11 day ascent to Everest Base Camp (EBC-early) (5300m), then after 6 weeks residence at Everest Base Camp (EBC-late), and finally on descent, over 5 days, to Kathmandu (KTM) (1300m) in May 2013.\n\nImages were obtained using a CytoCam-IDF video microscope (Braedius, Medical BV, Netherlands). This new device is based on the principle of Incident Dark Field (IDF) imaging, which uses polarized green light (wavelength 548nm) produced from LEDs to visualize, in real time, the sublingual microvasculature. Its high resolution imaging sensor (14 Mpixel) allows for a 50% increase in optical resolution (300 lines/mm) compared to previous Sidestream Dark Field imaging devices, and it generates a far larger field of view. With the participant lying in the supine position having rested for a minimum of 10 minutes, the CytoCam-IDF device’s probe was introduced into their mouth and placed on the mucosal surface of the inner lip. Once a suitable image was visualised on the screen of the CytoCam-IDF monitor, (Figure 1), 1 second of digital video footage was recorded. This process was conducted on all four lip quadrants (right upper lip, left upper lip, right lower lip, left lower lip), and at each quadrant four separate videos were acquired. Two trained investigators (EGK, PH) obtained all the data.\n\nTo determine capillary density, analysis was conducted offline by two independent investigators blinded to the participant identity, the testing conditions and the imaging site (EGK, JC). Using the company’s own video software (CytoCamTools V1, Braedius, Netherlands), a single still frame was projected on the computer screen and the number of capillary loops per image frame was counted manually. Partly visualised capillaries were included if the observer was assured that the vessel was a capillary due to its morphology. Subsequently, the mean capillary density was calculated from the four images obtained in each lip quadrant, and from these four results, the mean total lip density obtained. Capillary density was defined as the number of capillaries counted per field of view area (FOV), which equates to 1.79 mm2. The haemoglobin concentration ([Hb]) (Hemocue AB, Hemocue, Sweden) and haematocrit (Hct) (Sigma 1–14 microcentrifuge, Sigma, Germany) were obtained from whole blood samples, and peripheral arterial oxygen saturation (SpO2) measured (Nonin Onyx 9500, Nonin Medical Inc, Minnesota, USA) on the same days as microcirculatory imaging was performed.\n\nAs data were not normally distributed, they were described by median and interquartile range. Repeated sets of paired values were compared using Kruskall Wallis ANOVA, whilst comparisons of values between LON baseline and other sites was by Wilcoxon Signed Rank Test. Correlation between different variables was performed using Spearman’s rank correlation coefficient, and concordance between analysing investigators using intra-class correlation coefficient. All statistical analysis was undertaken on SPSS version 21 (SPSS Inc., Chicago, IL, USA), and a P value of <0.05 was taken to indicate statistical significance.\n\n\nResults\n\nCytoCam-IDF imaging was conducted on all ten individuals on the first two occasions, however only eight individuals had data captured on descent. No problems were encountered with the device or image acquisition. Mean laboratory barometric pressure and mean temperature for each location is shown in Table 1.\n\nChanges in labial capillary density are shown in Figure 2. Compared with LON (median 22.8 capillaries per field of view area (20.7–26.8)), capillary density was significantly increased at EBC-early 25.3 (24.5–30.6; p=0.021), EBC-late 32.5, (28.4–36.63; p=0.012), and on descent in KTM 31.0 (24.0–35.13; P=0.017). Between EBC-early and EBC-late, capillary density increased significantly (p=0.017), however there was no significant decline in density between EBC-late and KTM (p = 0.069).\n\nChanges in [Hb], Hct and SpO2 at each site are shown in Table 2. There was a significant increase in [Hb] between LON and EBC-early (p=0.007), and EBC-early and EBC-late (p=0.011), and a decrease between EBC-late and KTM (p=0.008). There was also a significant increase in Hct between LON and EBC-early (p=0.007), and EBC-late and KTM (p=0.012), but no significant change between EBC-early and EBC-late (p=0.191). Between the sites on ascent, the increase in vessel density demonstrated an inverse relationship with the SpO2, however at each altitude there was no correlation between vessel density and [Hb], Hct or SpO2.\n\nTo assess whether an image capture time of 1 second was indicative of that captured over longer periods of time, we obtained 30 seconds of footage from four individuals at two different locations. From this we randomly selected one frame per five seconds of footage, and counted the number of capillaries per field of view area. The values of these may be seen in Table 3, as too can the mean and standard deviations for each set of frames, the latter of which demonstrates a highest value of only 0.52 capillaries per field of view area.\n\ns = seconds; SD = standard deviation\n\n\nDiscussion\n\nThis study demonstrates for the first time, persistence of in vivo sublingual microvascular density increase on re-exposure to normoxia after a prolonged period of hypobaric hypoxia at high altitude. We utilised an infrequently used imaging and analysis technique that we had purposefully adapted to suit our needs, and found our data aligned with previously published work on blood vessel density at altitude10.\n\nUsing the data obtained from corresponding blood samples, it is possible to speculate on the adaptive processes occurring at each measurement point. As previously described, we observed a significant rise in [Hb] (14.5 g/dl to 16.4g/dl; p=0.007) and Hct (44% to 52%; p=007)) on ascent to altitude. Whilst this polycythaemia increases arterial oxygen content, blood viscosity also rises, altering its rheology. Under normal physiological conditions, a considerable proportion of the microcirculation is thought to be ‘unrecruited’, acting as a reservoir for times of increased metabolic needs15. As Hct rises, these reserve vessels are recruited, and microvascular density increases, along with functional capillary density15–18. Thus a secondary benefit to increased Hct is achieved; a reduction in the diffusion distance from capillaries to mitochondria. Importantly however, it should be noted that in normal capillary Hct is generally 50% less than systemic Hct owing to the streamlined blood flow in narrow capillaries19. The effect of hypoxia on this association is unknown. Whether or not neovascularisation had occurred on arrival at high altitude is difficult to say, although due to the short time between measurement points the chances of this being the case are low20.\n\nAfter 6 weeks spent at altitude, a further, and far greater, increase in vessel density was apparent; EBC-early 25.3 capillaries per field of view area, EBC-late 32.5 (p=0.017). Over the same time period, [Hb] had significantly risen, whilst Hct had not. As Hct is a more reliable indicator of viscosity between the two variables21, it seems unlikely that further recruitment of the microvasculature had occurred, yet it is plausible that neovascularisation had. Increased levels of vascular endothelial growth factor (VEGF) have been detected in subjects ascending to high altitude22; its role in angiogenesis perhaps explaining the observed rise in microvascular density10. Such adaptations lead to improved tissue oxygenation by a reduction of the inter-capillary distance, whilst maintaining a sufficiently low, and thus fluid Hct to permit flow of red blood cells in the microvasculature.\n\nOn descent to a lower altitude (KTM) there was no significant fall in microvascular density when compared to EBC-late (p = 0.069), however, a much greater number of vessels (36% increase) was evident when compared with baseline testing in LON. Whilst vessel density was thus unaltered on descent, over the same time point [Hb] and Hct values significantly declined (p=0.012). When compared to the original LON values, Hct on descent to KTM was significantly higher (44.0% and 50.0% respectively (p=0.011)), however, [Hb] was not (14.5 and 16.0g/dl (p=0.052)). Teasing apart the relative contributions of vessel recruitment and neovascularisation to the observed changes in sublingual microcirculatory density is challenging. Whilst the failure of vessel density to return to baseline after descent suggests some neovascularisation, neither [Hb] nor Hct had normalised at the time of the final readings so a raised blood viscosity could perhaps be maintaining a heightened level of capillary recruitment. A combination of the two processes would make sense as continually increasing [Hb] to improve oxygen delivery would eventually be counter productive. Indeed in Tibetans, who have been exposed to environmental hypoxia for many generations, there is a clear reduction in [Hb] compared to populations who have been exposed to these conditions for less time23–26. This suggests Tibetans utilize alternative long-term strategies for chronic adaptation to hypobaric hypoxia, ones that do not rely on maintaining a high [Hb]. It is plausible that one such means would be to increase their capillary density.\n\nThe use of the described methodology was also novel. A similar technique has been used twice previously; once in the assessment of coronary artery disease in diabetes13 and the other in a study investigating hypertension and rarefaction during treatment with Telatinib12. In these instances, data capture involved recording sublingual images for 1 minute13 or 30 seconds12 per quadrant, however we altered this time period by using an extremely short capture phase for data acquisition (< 1 second). Crucially, this allowed us to readily obtain snap shot images to reveal data about microvasculature density, whilst avoiding concerns surrounding probe and patient movement, in addition to issues relating to pressure artefact. Analysis was rapid, simple and reproducible; in this study it had an observer mean intra-class correlation coefficient of 0.91 (95% CI 0.84 – 0.96). Previously no difference in capillary density was observed in ten individuals between lip quadrants, and the reproducibility of the technique to determine capillary density was moderate to high with a coefficient of variation of 4.6%12. Of note, the technique does not allow assessment of microvascular flow, nor does it yield information on heterogeneity of microvascular blood flow, however, we propose it to be a robust method for the assessment of labial vessel density that could be conducted after only a short user training period.\n\nThe small number of participants used in this study could be considered a study limitation. As we were both employing a newly-adapted data acquisition technique, and using a novel device at altitude, no power calculation was performed. This therefore increases the risk of a type 2 error. Other limiting factors include the environmental considerations associated with high altitude research in a remote field environment. These include fluctuations in laboratory temperature, humidity (Table 1) and participant hydration status, all factors that may alter microvascular blood flow and density. Attempts were made to limit these potential confounding factors by performing CytoCam-IDF imaging at the same time of day in heated purpose-built laboratories, and encouraging participants to maintain a good state of hydration. Previous studies at altitude have also raised concerns over the development of tissue oedema10,11 that can occur on ascent to altitude27. Whilst this could potentially reduce image quality and lead to false measurements of flow and density, our IDF camera provided us with a depth of focus reading, thus allowing us to confirm that we were recording at the same depth under the tongue on each time point. Finally, we have discussed alterations in capillary or vessel density. It is important however to clarify this nomenclature. IDF imaging cannot image blood vessels directly but rather uses the fact that polarized green light is optimally absorbed by red blood cells within the microvasculature regardless of oxygenation status. Absorption of light by haemoglobin, but not by surrounding tissues, therefore creates a distinct contrast of dark and light colour respectively, and red blood cells moving through the mucosal microcirculation thus appear as dark globules moving along the axis of flow. All vessels visualized are therefore only seen if they contain erythrocytes. The variables measured by IDF imaging (and its precursor SDF imaging) include a measure of total vessel density (TVD) and perfused vessel density (PVD)28. A distinction is made between the two depending on the speed of red blood cell flow within the observed vessels. TVD includes vessels which contain erythrocytes flowing at any velocity (or even at standstill), whilst PVD only includes vessels with continuously moving erythrocytes. As we cannot measure erythrocyte velocity with this simplified method, our observations therefore describe the TVD.\n\n\nConclusions\n\nThis study demonstrated an increase in sublingual microvascular vessel density on early and sustained exposure to hypobaric hypoxia; and, for the first time, that no significant change in vessel density occurred on immediate descent. The technique used to capture the images provided a rapid and reliable means for assessing changes in vessel density, and could be applied in future studies of microcirculatory vessel density. Further research in this area may allow a more complete comprehension of the multidimensional response to sustained hypoxia that occurs during pathophysiological situations.\n\n\nData availability\n\nF1000Research: Dataset 1. IDF values, 10.5256/f1000research.7649.d13402129\n\nF1000Research: Dataset 2. Physiological values, 10.5256/f1000research.7649.d13402230",
"appendix": "Author contributions\n\n\n\nE G-K: design of study, collection of data, analysis of data, writing manuscript\n\nJC: collection of data, analysis of data, writing manuscript\n\nPH: analysis of data, writing manuscript\n\nMG: design of study, writing manuscript\n\nCI: design of study, writing manuscript\n\nDM: design of study, analysis of data, writing manuscript\n\nAll authors have seen and agreed to the final content of the manuscript\n\n\nCompeting interests\n\n\n\nBraedius Medical, a company owned by a relative of Can Ince, has developed and designed a hand held microscope called CytoCam-IDF imaging. Can Ince has no financial relation with Braedius Medical of any sort; he never owned shares, or received consultancy or speaker fees from Braedius Medical.\n\n\nGrant information\n\nXtreme Everest 2 was supported by the Royal Free Hospital NHS Trust Charity, the Special Trustees of University College London Hospital NHS Foundation Trust, the Southampton University Hospital Charity, the UCL Institute of Sports Exercise and Health, The London Clinic, University College London, University of Southampton, Duke University Medical School, the United Kingdom Intensive Care Society, the National Institute of Academic Anaesthesia, the Rhinology and Laryngology Research Fund, The Physiological Society, Smiths Medical, Deltex Medical, Atlantic Customer Solutions and the Xtreme Everest 2 volunteer participants who trekked to Everest Base Camp.\n\nSome of this work was undertaken at University College London Hospital-University College London Biomedical Research Centre, which received a proportion of funding from the United Kingdom Department of Health’s National Institute for Health Research Biomedical Research Centers funding scheme. Some of this work was undertaken at University Hospital Southampton-University of Southampton Respiratory Biomedical Research Unit, which received a proportion of funding from the United Kingdom Department of Health’s National Institute for Health Research Biomedical Research Units funding scheme.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nXtreme Everest 2 is a research project coordinated by the Caudwell Xtreme Everest Hypoxia Research Consortium, collaboration between the UCL Centre for Altitude, Space, and Extreme Environment Medicine, the Centre for Human Integrative Physiology at the University of Southampton and the Duke University Medical Centre. Membership, roles and responsibilities of the Xtreme Everest 2 Research Group can be found at www.xtreme-everest.co.uk/team.\n\n\nReferences\n\nWagner PD: Reduced maximal cardiac output at altitude--mechanisms and significance. Respir Physiol. 2000; 120(1): 1–11. PubMed Abstract | Publisher Full Text\n\nStoneham MD, Pethybridge RJ: Acclimatization to altitude: effects on arterial oxygen saturation and pulse rate during prolonged exercise at altitude. J R Nav Med Serv. 1993; 79(1): 3–9. PubMed Abstract\n\nPugh LG: Blood volume and haemoglobin concentration at altitudes above 18,000 ft. (5500 m). J Physiol. 1964; 170(2): 344–354. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevett DZ, Fernandez BO, Riley HL, et al.: The role of nitrogen oxides in human adaptation to hypoxia. Sci Rep. 2011; 1: 109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHochachka PW, Stanley C, Merkt J, et al.: Metabolic meaning of elevated levels of oxidative enzymes in high altitude adapted animals: an interpretive hypothesis. Respir Physiol. 1983; 52(3): 303–313. PubMed Abstract | Publisher Full Text\n\nHoppeler H, Kleinert E, Schlegel C, et al.: Morphological adaptations of human skeletal muscle to chronic hypoxia. Int J Sports Med. 1990; 11(Suppl 1): S3–S9. PubMed Abstract | Publisher Full Text\n\nMacDougall JD, Green HJ, Sutton JR, et al.: Operation Everest II: structural adaptations in skeletal muscle in response to extreme simulated altitude. Acta Physiol Scand. 1991; 142(3): 421–427. PubMed Abstract | Publisher Full Text\n\nLundby C, Pilegaard H, Andersen JL, et al.: Acclimatization to 4100 m does not change capillary density or mRNA expression of potential angiogenesis regulatory factors in human skeletal muscle. J Exp Biol. 2004; 207(pt 22): 3865–3871. PubMed Abstract | Publisher Full Text\n\nMizuno M, Savard GK, Areskog NH, et al.: Skeletal muscle adaptations to prolonged exposure to extreme altitude: a role of physical activity? High Alt Med Biol. 2008; 9(4): 311–317. PubMed Abstract | Publisher Full Text\n\nMartin DS, Goedhart P, Vercueil A, et al.: Changes in sublingual microcirculatory flow index and vessel density on ascent to altitude. Exp Physiol. 2010; 95(8): 880–891. PubMed Abstract | Publisher Full Text\n\nMartin DS, Ince C, Goedhart P, et al.: Abnormal blood flow in the sublingual microcirculation at high altitude. Eur J Appl Physiol. 2009; 106(3): 473–478. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSteeghs N, Gelderblom H, Roodt JO, et al.: Hypertension and rarefaction during treatment with telatinib, a small molecule angiogenesis inhibitor. Clin Cancer Res. 2008; 14(11): 3470–3476. PubMed Abstract | Publisher Full Text\n\nDjaberi R, Schuijf JD, de Koning EJ, et al.: Non-invasive assessment of microcirculation by sidestream dark field imaging as a marker of coronary artery disease in diabetes. Diab Vasc Dis Res. 2013; 10(2): 123–134. PubMed Abstract | Publisher Full Text\n\nMartin DS, Gilbert-Kawai E, Levett DZh, et al.: Xtreme Everest 2: unlocking the secrets of the Sherpa phenotype? Extrem Physiol Med. 2013; 2(1): 30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParthasarathi K, Lipowsky HH: Capillary recruitment in response to tissue hypoxia and its dependence on red blood cell deformability. Am J Physiol. 1999; 277(6 Pt 2): H2145–H2157. PubMed Abstract\n\nGenzel-Boroviczény O, Christ F, Glas V: Blood transfusion increases functional capillary density in the skin of anemic preterm infants. Pediatr Res. 2004; 56(5): 751–755. PubMed Abstract | Publisher Full Text\n\nSakr Y, Chierego M, Piagnerelli M, et al.: Microvascular response to red blood cell transfusion in patients with severe sepsis. Crit Care Med. 2007; 35(7): 1639–1644. PubMed Abstract | Publisher Full Text\n\nYuruk K, Almac E, Bezemer R, et al.: Blood transfusions recruit the microcirculation during cardiac surgery. Transfusion. 2011; 51(5): 961–967. PubMed Abstract | Publisher Full Text\n\nDuling BR, Klitzman B: Local control of microvascular function: role in tissue oxygen supply. Annu Rev Physiol. 1980; 42: 373–382. PubMed Abstract | Publisher Full Text\n\nLindeboom JA, Mathura KR, Aartman IH, et al.: Influence of the application of platelet-enriched plasma in oral mucosal wound healing. Clin Oral Implants Res. 2007; 18(1): 133–139. PubMed Abstract | Publisher Full Text\n\nPries AR, Secomb TW, Gaehtgens P: Biophysical aspects of blood flow in the microvasculature. Cardiovasc Res. 1996; 32(4): 654–667. PubMed Abstract | Publisher Full Text\n\nWalter R, Maggiorini M, Scherrer U, et al.: Effects of high-altitude exposure on vascular endothelial growth factor levels in man. Eur J Appl Physiol. 2001; 85(1–2): 113–117. PubMed Abstract | Publisher Full Text\n\nAdams WH, Strang LJ: Hemoglobin levels in persons of tibetan ancestry living at high altitude. Proc Soc Exp Biol Med. 1975; 149(4): 1036–1039. PubMed Abstract | Publisher Full Text\n\nBeall CM, Reichsman AB: Hemoglobin levels in a Himalayan high altitude population. Am J Phys Anthropol. 1984; 63(3): 301–306. PubMed Abstract | Publisher Full Text\n\nSamaja M, Veicsteinas A, Cerretelli P: Oxygen affinity of blood in altitude Sherpas. J Appl Physiol Respir Environ Exerc Physiol. 1979; 47(2): 337–341. PubMed Abstract\n\nWu T, Wang X, Wei C, et al.: Hemoglobin levels in Qinghai-Tibet: different effects of gender for Tibetans vs. Han. J Appl Physiol (1985). 2005; 98(2): 598–604. PubMed Abstract | Publisher Full Text\n\nMaggiorini M, Bühler B, Walter M, et al.: Prevalence of acute mountain sickness in the Swiss Alps. BMJ. 1990; 301(6756): 853–855. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Backer D, Hollenberg S, Boerma C, et al.: How to evaluate the microcirculation: report of a round table conference. Crit Care. 2007; 11(5): R101. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGilbert-Kawai E, Coppel J, Phillip H, et al.: Dataset 1 in: Changes in labial capillary density on ascent to and descent from high altitude. F1000Research. 2016. Data Source\n\nGilbert-Kawai E, Coppel J, Phillip H, et al.: Dataset 2 in: Changes in labial capillary density on ascent to and descent from high altitude. F1000Research. 2016. Data Source"
}
|
[
{
"id": "16269",
"date": "14 Sep 2016",
"name": "Sam Hutchings",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nEdward Gilbert-Kawai and colleagues involved in the Xtreme Everest 2 study report further work on the effects of altitude and hypoxaemia on the microcirculation. They describe a technique for rapidly measuring microcirculatory capillary density, using a method of counting the number of capillary loops in the buccal mucosa of healthy volunteers at sea level, progressive altitude acclimatisation and descent. They also report systemic haemoglobin and haematocrit values in an attempt to elucidate whether changes in microcirculatory parameters are paralleled by changes in blood viscosity. Their results show, in line with previous work, that microcirculatory vessel density increases after prolonged exposure to hypoxic conditions. However, they also show that there is some persistence of this effect on partial descent. The authors pose a question as to whether this increased vessel density is due to neo-vascularisation or recruitment of existing capillaries secondary to an increase in haematocrit.\n\nThis is generally a scientifically sound paper and a well written manuscript. I have several comments for the authors:\n\nMajor Comments\nIt strikes me that the authors are trying to achieve two things in this paper. Firstly, to confirm that there is microcirculatory adaptation to altitude and hypoxaemia and to provide more information regarding the mechanisms surrounding this and secondly to describe a method of rapidly assessing vessel density. Both subjects are not quite covered in enough detail to do either justice.\n\nWith regard to the assessment method, this has been described previously in both buccal mucosa but also in other areas such as the nail bed. It has the advantage of being quick to perform but only provides information on one variable – number of capillary loops. The authors state and assume that this measure equates to Total Vessel Density (TVD) but it would be very useful to confirm this by actually measuring capillary TVD using sublingual IDF video capture and analysis. More importantly, the described method provides no measure of microcirculatory flow. This is crucial when discussing the impact of capillary haematocrit and viscosity on microcirculatory performance. Personally I think the paper would have been improved had the investigators used sublingual images and analyzed these to produce data on a fuller range of flow and density parameters.\n\nAlthough the authors give two possible mechanisms for the observed increase in capillary density, I would be interested to know if these changes could possibly be accounted for by an increase in systemic haemodynamic parameters, such as cardiac output. Unless there is a loss of haemodynamic coherence one would expect that micro would follow macro. Measurement of these values using a non invasive technique such as trans thoracic echocardiography would have been interesting and added value to the paper.\n\nThe authors discuss that systemic haematocrit and haemoglobin do not necessarily equate with capillary haemoglobin. This is a crucial point and one often observes that “shocked” capillaries have a visibly lower number of erythrocytes even in the presence of a reasonably preserved systemic Hb level, especially post resuscitation. Being able to quantify this difference would allow us to answer crucial questions but is not possible with current analysis methods using this technology.\n\nI would be very interested to know how long this microcirculatory adaptation lasts. Did the authors consider repeating the measures after return to sea level and after a more prolonged period – e.g. one month?\n\nMinor Comments\nSouthampton is spelt incorrectly in several places within the list of authors\n\nCiting a validation paper of the IDF camera (e.g. Hutchings S, Watts S, Kirkman E. The Cytocam video microscope. A new method for visualising the microcirculation using Incident Dark Field technology. Clin Hemorheol Microcirc. IOS Press; 2015 Oct 16;62(3):261–71) would be useful as the authors have stated that it is a new device.",
"responses": []
},
{
"id": "15955",
"date": "28 Sep 2016",
"name": "Matthias Hilly",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report on the use of a relatively novel method to determine capillary density during acute and chronic exposure to hypobaric hypoxia. Previously reported results of an increase in capillary density during acclimatization are confirmed, but it is for the first time demonstrated that upon immediate descent some changes persist.\nOverall, this is a well-conducted study that adds to the current knowledge base of microcirculatory changes during adaptation to hypobaric hypoxia, and provides an innovative approach to vessel density measurement.\n\nSpecific comments:\nIntroduction\nIn the introduction the absence of neovascularization in muscle samples during adaptation to high altitude is discussed. It is however also known that mitochondrial density increases during this process (PMID 26339730). The authors may want to include this in the discussion about capillary-to-fibre ratio, the mitochondria arguably representing the circulation's endpoint.\nMethods\nThe KTM time point seems difficult to me for use as a “post-hypoxia” measurement, since it is not comparable to the LON baseline. This is not a severe limitation since “immediate descent” was one of the study endpoints rather that a comparison to the initial (LON) baseline. I still suggest mentioning this in the limitations.\nMicrocirculation capture time is very short (1 sec). I am not sure if the analysis in four subjects given in table 3 is sufficient support for the use of much shorter capture intervals than recommended by current consensus guidelines. This could be discussed in more detail in the limitations.\nHb and Hematocrit values, as well as SpO2 were determined “on the same day as microcirculatory imaging was performed”. Given the influence of hydration status changes throughout a day, especially in a mountain setting, either specify the time frame more clearly or mention this time relation in the limitations.\nResults\nFigure 2: It is unclear to me what the results of the Kruskal-Wallis ANOVA described in the methods section were for vessel density measurements, consider mentioning these in the results section. Further, are the P values given in the text references to Wilcoxon Signed Rank tests as mentioned in the methods section? Consider describing these tests as post-hoc analysis for the non-parametric ANOVA. And if so, was a correction for multiple testing applied (Bonferroni, or one of the newer algorithms)? Please describe this more clearly.\nIt is mentioned that at each altitude there was no correlation between vessel density and Hb, Hct or SpO2. In the methods section it is stated that Spearman's rank correlation coefficient was calculated. Consider adding correlation coefficients and P values to the results section to provide the readers some insight into the data.\nIn addition to Hb, Hct and SpO2 given in Table 2, it would be of interest to know more about the subjects' physiologic parameters, such as blood pressure, cardiac output and even RV/RA – including a discussion about the changes in macrocirculation in relation to the changes in microcirculation.\nDiscussion\nGiven borderline significance (p=0.069) for a decrease in vessel density from EBC-late to KTM and the lack of a power calculation, as well as the lack of real comparability to the initial (LON) baseline it is somewhat difficult to argue that there really was no change between EBC-late and KTM. This is discussed in the discussion to some extent, but it should be stated more clearly that a verification of these results is needed in the future.\nThe method for vessel density estimation used in this study differs from the effective calculation of total vessel density (TVD), being the total length of small vessels divided by the surface area covered in the recording, that is recommended by current consensus guidelines. Please discuss the relationship between vessel density as measured in this study with vessel density as assessed by TVD, and potential confounders.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2107
|
https://f1000research.com/articles/5-1745/v1
|
19 Jul 16
|
{
"type": "Software Tool Article",
"title": "Contextual Hub Analysis Tool (CHAT): A Cytoscape app for identifying contextually relevant hubs in biological networks",
"authors": [
"Tanja Muetze",
"Ivan H. Goenawan",
"Heather L. Wiencko",
"Manuel Bernal-Llinares",
"Kenneth Bryan",
"David J. Lynn",
"Tanja Muetze",
"Ivan H. Goenawan",
"Heather L. Wiencko",
"Manuel Bernal-Llinares",
"Kenneth Bryan"
],
"abstract": "Highly connected nodes (hubs) in biological networks are topologically important to the structure of the network and have also been shown to be preferentially associated with a range of phenotypes of interest. The relative importance of a hub node, however, can change depending on the biological context. Here, we report a Cytoscape app, the Contextual Hub Analysis Tool (CHAT), which enables users to easily construct and visualize a network of interactions from a gene list of interest, integrate contextual information, such as gene expression data, and identify hub nodes that are more highly connected to contextual nodes (e.g. genes that are differentially expressed) than expected by chance. In a case study, we use CHAT to construct a network of genes that are differentially expressed in Dengue fever, a viral infection. CHAT was used to identify and compare contextual and degree-based hubs in this network. The top 20 degree-based hubs were enriched in pathways related to the cell cycle and cancer, which is likely due to the fact that proteins involved in these processes tend to be highly connected in general. In comparison, the top 20 contextual hubs were enriched in pathways commonly observed in a viral infection including pathways related to the immune response to viral infection. This analysis shows that such contextual hubs are considerably more biologically relevant than degree-based hubs and that analyses which rely on the identification of hubs solely based on their connectivity may be biased towards nodes that are highly connected in general rather than in the specific context of interest.\n\nAvailability: CHAT is available for Cytoscape 3.0+ and can be installed via the Cytoscape App Store (http://apps.cytoscape.org/apps/chat).",
"keywords": [
"Network analysis",
"hypergeometric test",
"hubs",
"gene expression data",
"contextual hub analysis",
"CHAT"
],
"content": "Introduction\n\nNetwork analysis has emerged as a powerful approach to elucidate biological and disease processes1. Biological networks (and many other types of networks) have been shown to have a power law distribution of node connectivity, with most nodes having few connections and a few nodes being highly connected2. The identification of such highly connected nodes, termed hubs, is often of interest as hubs have been shown to be topologically and functionally important. The deletion of genes encoding hub proteins, for example, has been shown to correlate with lethality in yeast (the centrality-lethality rule)3. Hubs have also been found to be preferentially targeted by both bacterial and viral pathogens4 and may be master regulators of biological processes5. Biological networks, such as the human interactome, however, are not static entities6, and the extent to which a node acts as a hub can change depending on the biological context e.g. the network present in a specific cell type at a particular point in time7,8. Integrating contextual information, such as gene expression data, with standard network analysis can provide insight into what are the most relevant network features in a particular study or context9–11.\n\nCytoscape has a number of applications to identify hubs in networks including cytoHubba12, APID2Net13, PinnacleZ14, NetworkAnalyzer15,16 and CentiScaPe17, however, only the latter two are compatible with Cytoscape 3+. All of the applications available to date identify hubs based on node connectivity (degree) in a network of interest. To construct a network, users frequently query interaction databases to identify the interactors of a list of genes of interest, e.g. differentially expressed genes, and then identify the high degree nodes in this network. This approach to constructing a network is useful because it identifies a more fully connected network for analysis than would be the case if one restricted interactions to only those that occur between nodes in the gene list. Analysis of these networks can, for example, identify subnetworks that are enriched in (but do not exclusively consist of) differentially expressed genes, or identify non-differentially expressed nodes that are topologically important in the network, both of which would otherwise not be identified. Identifying hubs in these networks, however, is biased towards identifying nodes that are highly connected in general such as promiscuous, ubiquitous or well-studied nodes, because nodes with many interactions in the query database have a higher probability of being included in the network by chance alone. Analysis of these degree-based hubs, for example identifying what biological processes or pathways these nodes are enriched in, tells us little about the experimental context of interest and more about the properties of highly connected nodes in general. A more appropriate analysis is to determine which nodes interact with relevant nodes in the network (which we term contextual nodes) more than is statistically expected.\n\nHere, we introduce the Contextual Hub Analysis Tool (CHAT), a Cytoscape App that identifies hub nodes that interact with more \"contextual\" nodes (e.g. differentially expressed genes) than statistically expected in networks integrated with user-supplied contextual data (e.g. gene expression data). We term these nodes contextual hubs. We show that such contextual hubs are considerably more relevant than degree-based hubs to the specific experimental context under investigation. As such, these nodes are promising candidates for further functional validation studies and potentially represent important points in the network for drug targeting.\n\n\nMethods\n\nCHAT was written in Java 8 as an Open Services Gateway Initiative (OSGi) bundle for Cytoscape 3.0+18. It adds a “CHAT” option in the “Apps” menu that launches a popup window, which allows users to adjust different network initialization parameters. CHAT prompts users to input a list of gene identifiers (the supported ID types are dependent on the database selected by the user) and any associated contextual data, e.g. gene expression data associated with the genes. The OK button triggers Cytoscape’s TaskManager to run a task that initiates the network construction and adds a tab to the results panel that provides functionality to further modify and analyze the network. To create the network, CHAT finds all the first neighbor interactors of the user-provided genes (or their encoded products). Interaction data is retrieved from one of the databases included in the PSICQUIC registry19, which the user can select. Once the network has been constructed, CHAT performs a hypergeometric test on each node in the network to identify nodes that interact with contextual nodes more than expected by chance. The probability that a given hub has k or more contextual interactors among its n interactors is given by the hypergeometric distribution:\n\np(X≥k)=∑x=kn(Kx)(N−Kn−x)(Nn)\n\nWhere N is the number of genes with at least one interaction in the database queried and K is the number of contextually relevant nodes provided by the user (with at least one interaction in the database queried). Overrepresentation analysis heavily depends on the choice of background dataset for the determination of N. To estimate the background frequency K/N, CHAT provides access to interaction data from databases available in the PSICQUIC registry. Databases with less than 10,000 interactions are excluded. The number of genes in the user-selected database that have at least one interaction (of the specified type) in which both interactors match the user-selected criteria for constructing the network (species, interaction type and ID type) determine the node population size N. Self-interactions are disregarded. P-values calculated by CHAT are automatically corrected for multiple testing using the Benjamini-Hochberg procedure20, a method widely used in bioinformatics to avoid high false discovery rates. The Bonferroni approach is widely considered to be too strict21.\n\nA right click on a node brings up an option to activate the “Node Analyzer” mode, which allows the user to analyze the connectivity pattern of individual hubs of interest. Using this function will display the node analyzer table on the results panel and all nodes except the selected node and its interactors will be hidden in the network visualization. The execution time of CHAT varies between a few seconds and a few minutes based on the number of user-supplied (contextual) genes, the size of the chosen database and its connection speed as well as the user-selected network layout. These factors also influence memory consumption.\n\nThe identification of the top contextual hubs consists of three primary steps: 1) input of a user-supplied gene list and contextual data, 2) network construction and statistical analysis to identify nodes that preferentially interact with contextual nodes and 3) visualization of the top contextual hubs and their interactions and comparison to the top degree-based hubs. To construct a network using CHAT, the user must provide a list of gene identifiers and associated numerical or categorical attributes in the text box in tab-delimited format, or upload the data as a csv or tab-delimited file via the upload button (Figure 1) (.csv or .txt file types). The user can then specify which genes in the uploaded list are contextually important based on the user-provided contextual data (e.g. genes with > 2 fold-change in expression). The user then selects one of the databases in the PSICQUIC registry to query, and specifies the relevant species, ID type and interaction type for the query. The user can then choose to visualize the network using any of the layout algorithms available in Cytoscape. Clicking the OK button creates the network and a new tab in the results panel, which allows the user to visualize the network and to analyze the results further (Figure 2). The results panel is split into several parts. In the first part, the parameters used to generate the network (database, species, id type and interaction type(s)) are displayed. The second panel allows the user to compare the top contextual hubs and the top degree-based hubs at the click of a button. By default, node size and node color are proportional to the node’s corrected p-value calculated by CHAT, such that the smaller the p-value (i.e. more statistically significant), the larger the node size and the darker the red coloring of the node. The user can customize the color scheme, however. In contrast, if the users selects “Show degree hubs”, the visualization changes and the node size and coloring will now be proportional to each node’s degree in the selected database. By default, CHAT displays the top 20 contextual hubs but the user can adjust this by using the slider provided. To investigate a single node in detail the user can employ CHAT’s “Node Analyzer” by right clicking on a node. This will limit the network view to show only the selected node and its interactors and will display a table at the bottom of the results panel tab with information on the node’s name, p-value and its interactors.\n\nTo construct a network using CHAT, the user provides a list of gene identifiers and associated numerical or categorical attributes relevant in the context of interest.\n\nCHAT provides a number of options to customize the network visualization.\n\n\nUse case\n\nAs a demonstration of its potential utility and as validation, CHAT was used to construct a network using a dataset of 462 genes that have been reported to be up-regulated during Dengue fever, a mosquito-borne viral infection22 (Ensembl gene IDs for these 462 genes are provided in Dataset 1). These 462 genes represent the contextual data for this case study. CHAT was used to construct a network of these genes and their first neighbor interactors using interaction data that was sourced from InnateDB23 via the PSICQUIC web service (InnateDB-All). A network of 4,910 nodes was generated. CHAT was then used to identify the top 20 conventional hub nodes (based solely on degree) and the top 20 contextual hub nodes in the network (Figure 3). No nodes were in common in the two top 20 lists. InnateDB pathway analysis23 revealed that the top 20 degree-based hubs were enriched in pathways related to the cell cycle and cancer (Supplementary Table 1), which is likely due to the fact that proteins involved in these processes tend to be highly connected in general. In comparison to degree-based hubs, the top 20 contextual hubs were statistically enriched in pathways related to the immune response to viral infection, such as the interferon signaling pathway; the Retinoic acid inducible gene-I (RIG-I) pathway; the Toll-like receptor (TLR) pathway; and the Janus kinase (JAK) - Signal Transducer and Activator of Transcription (STAT) pathway (Supplementary Table 2). All of these pathways have been shown to play key roles in the host response to Dengue infection24,25. Indeed, many of the top 20 contextual hubs (but not degree-based hubs) were well-known transcription factors involved in the host interferon response including STAT1, STAT2 and the interferon regulatory factors (IRFs); IRF1, 3, 8 and 9, which is a key cellular response to viral infection including Dengue26,27. Another gene identified in the contextual hub analysis but not the degree-based analysis was interferon-stimulated gene 15 (ISG15). Cells in which ISG15 has been silenced have been shown to have significantly higher Dengue viral loads28. The results of the pathway analysis were reinforced by a Gene Ontology analysis using innatedb.com23, which identified terms including cytokine-mediated signaling pathway, type I interferon signaling pathway, and innate immune response among the top 10 enriched terms (FDR < 0.05) for the contextual hubs but not the degree-based hubs (Supplementary Table 3 and Supplementary Table 4).\n\nA CHAT network visualization comparing contextual hubs (A) to degree-based hubs (B) in a network constructed using InnateDB23.\n\n\nConclusion\n\nThrough the integration of contextual information, such as gene or protein expression, contextual hub analysis as implemented in CHAT can identify context-specific hubs more relevant to the biological context under study, such as disease, treatment or cellular state. As shown in the above case study, these hubs are of more functional relevance than genes found through analysis based on degree only. Given the current emphasis on the importance of considering the network model of biological pathways and the ever-increasing abundance of high-throughput data, CHAT provides a valuable addition to the biologists’ computational toolkit in using a network-based approach to help prioritize genes of interest for further investigation or drug discovery. In the future, CHAT can be extended to include the contextual analysis of other network features such as network bottlenecks.\n\n\nData availability\n\nF1000Research: Dataset 1. Use case data: 462 genes that have been reported to be up-regulated during Dengue fever infection, 10.5256/f1000research.9118.d12812629\n\n\nSoftware availability\n\nSoftware available from: http://apps.cytoscape.org/apps/chat\n\nLatest source code: https://bitbucket.org/dynetteam/chat\n\nArchived source code at time of publication: http://www.dx.doi.org/10.5281/zenodo.5649630\n\nManual/Tutorial: https://bitbucket.org/dynetteam/chat/downloads\n\nLicense: Lesser GNU Public License 3.0",
"appendix": "Author contributions\n\n\n\nTM and IHG jointly developed the App under the supervision of KB and DJL. TM and DJL wrote the paper with contributions from the other authors. HLW developed an earlier unpublished Python version of CHAT that inspired the development of this App. MBL provided computational and systems support for the project. DJL conceived of the idea, supervised the App’s development and co-wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe research leading to these results received funding from the European Union Seventh Framework Programme (FP7/2007-2013) PRIMES project under grant agreement number FP7-HEALTH-2011-278568. The Lynn Group is also supported by EMBL Australia.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nClick here to access data. Supplementary Table 1. Pathway analysis of the top 20 degree-based hubs\n\nClick here to access data. Supplementary Table 2. Pathway analysis of the top 20 contextual hubs.\n\nClick here to access data. Supplementary Table 3. Gene ontology terms overrepresented among the top 20 contextual hubs in the Dengue fever network.\n\nClick here to access data. Supplementary Table 4. Gene ontology terms overrepresented among the top 20 degree-based hubs in the Dengue fever network.\n\n\nReferences\n\nBarabasi AL, Gulbahce N, Loscalzo J: Network medicine: a network-based approach to human disease. Nat Rev Genet. 2011; 12(1): 56–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarabasi AL, Oltvai ZN: Network biology: understanding the cell’s functional organization. Nat Rev Genet. 2004; 5(2): 101–113. PubMed Abstract | Publisher Full Text\n\nJeong H, Mason SP, Barabási AL, et al.: Lethality and centrality in protein networks. Nature. 2001; 411(6833): 41–42. PubMed Abstract | Publisher Full Text\n\nDyer MD, Murali TM, Sobral BW: The landscape of human proteins interacting with viruses and other pathogens. PLoS Pathog. 2008; 4(2): e32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBorneman AR, Leigh-Bell JA, Yu H, et al.: Target hub proteins serve as master regulators of development in yeast. Genes Dev. 2006; 20(4): 435–448. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrzytycka TM, Singh M, Slonim DK: Toward the dynamic interactome: It’s about time. Brief Bioinform. 2010; 11(1): 15–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRachlin J, Cohen DD, Cantor C, et al.: Biological context networks: a mosaic view of the interactome. Mol Syst Biol. 2006; 2: 66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgarwal S, Deane CM, Porter MA, et al.: Revisiting date and party hubs: Novel approaches to role assignment in protein interaction networks. PLoS Comput Biol. 2010; 6(6): e1000817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGao S, Wang X: Identification of highly synchronized subnetworks from gene expression data. BMC Bioinformatics. 2013; 14(Suppl 9): S5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZinman GE, Naiman S, O'Dee DM, et al.: ModuleBlast: identifying activated sub-networks within and across species. Nucleic Acids Res. 2015; 43(3): e20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoul J, Hardingham TE, Boot-Handford RP, et al.: PhenomeExpress: a refined network analysis of expression datasets by inclusion of known disease phenotypes. Sci Rep. 2015; 5: 8117. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChin CH, Chen SH, Wu HH, et al.: cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol. 2014; 8(Suppl 4): S11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHernandez-Toro J, Prieto C, De Las Rivas J: APID2NET: Unified interactome graphic analyzer. Bioinformatics. 2007; 23(18): 2495–2497. PubMed Abstract | Publisher Full Text\n\nChuang HY, Lee E, Liu YT, et al.: Network-based classification of breast cancer metastasis. Mol Syst Biol. 2007; 3: 140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAssenov Y, Ramírez F, Schelhorn SE, et al.: Computing topological parameters of biological networks. Bioinformatics. 2008; 24(2): 282–284. PubMed Abstract | Publisher Full Text\n\nDoncheva NT, Assenov Y, Domingues FS, et al.: Topological analysis and interactive visualization of biological networks and protein structures. Nat Protoc. 2012; 7(4): 670–85. PubMed Abstract | Publisher Full Text\n\nScardoni G, Tosadori G, Faizan M, et al.: Biological network analysis with CentiScaPe: centralities and experimental dataset integration [version 2; referees: 2 approved]. F1000Research. 2014; 3: 139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: A software Environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAranda B, Blankenburg H, Kerrien S, et al.: PSICQUIC and PSISCORE: accessing and scoring molecular interactions. Nat Methods. 2011; 8(7): 528–529. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenjamini Y, Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B. 1995; 57(1): 289–300. Reference Source\n\nNoble WS: How does multiple testing correction work? Nat Biotechnol. 2009; 27(12): 1135–1137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoang LT, Lynn DJ, Henn M, et al.: The early whole-blood transcriptional signature of dengue virus and features associated with progression to dengue shock syndrome in Vietnamese children and young adults. J Virol. 2010; 84(24): 12982–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBreuer K, Foroushani AK, Laird MR, et al.: InnateDB: systems biology of innate immunity and beyond--recent updates and continuing curation. Nucleic Acids Res. 2013; 41(Database issue): D1228–1233. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNasirudeen AM, Wong HH, Thien P, et al.: RIG-I, MDA5 and TLR3 synergistically play an important role in restriction of dengue virus infection. PLoS Negl Trop Dis. 2011; 5(1): e926. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSouza-Neto JA, Sim S, Dimopoulos G: An evolutionary conserved function of the JAK-STAT pathway in anti-dengue defense. Proc Natl Acad Sci U S A. 2009; 106(42): 17841–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe La Cruz Hernández SI, Puerta-Guardo H, Flores-Aguilar H, et al.: A strong interferon response correlates with a milder dengue clinical condition. J Clin Virol. 2014; 60(3): 196–199. PubMed Abstract | Publisher Full Text\n\nMorrison J, García-Sastre A: STAT2 signaling and dengue virus infection. JAKSTAT. 2014; 3(1): e27715. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDai J, Pan W, Wang P: ISG15 facilitates cellular antiviral response to dengue and west nile virus infection in vitro. Virol J. 2011; 8: 468. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuetze T, Goenawan IH, Wiencko HL, et al.: Dataset 1 in: Contextual Hub Analysis Tool (CHAT): A Cytoscape app for identifying contextually relevant hubs in biological networks. F1000Research. 2016. Data Source\n\nMuetze T, Goenawan IH, Wiencko HL, et al.: Contextual Hub Analysis Tool (CHAT): A Cytoscape app for identifying contextually relevant hubs in biological networks. Zenodo. 2016. Data Source"
}
|
[
{
"id": "15065",
"date": "09 Aug 2016",
"name": "Sandra Orchard",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well written technical paper, clearly outlining a new Cytoscape App in terms that would make it easy for a new user, with some familiarity with Cytoscape, to download, install and use. The ability to generate contextual hubs is currently not possible with existing Cytoscape Apps, so this is a valuable addition to the collection. A couple of queries and some minor points for correction:\nThe application searches for first-neighbour interactions of molecules in the list presented to it. It did not appear to search for interaction between members of the list, which should not affect the contextual nodes selection much, but will alter the degree-based hubs. This should be commented on, or the documentation made clearer if we are incorrect with this observation. To bypass this problem and make the user more aware of this limitation, the tool should be able to provide more control over how the network is constructed, for example providing the option to exclude first neighbours.\n\nCan this application be made to work with an existing network?\n\nThe text is entirely gene-centric and may leave an inexperienced use under the impression is is only usable for gene-expression data whereas it is equally useful for the analysis of proteomic data and works with UniProtKB identifiers. Whilst I realise this is apparent to anyone who downloads the app, it may well be worth adding a sentence to both the Summary or Introduction of this paper, and also the description in the App store just to make this very clear to naive users.\n\nIt may also be worth adding the reference to the 2013 PSICQUIC paper as well as the original as I personally find it more informative and again, may be helpful to the inexperienced user.",
"responses": [
{
"c_id": "2151",
"date": "30 Aug 2016",
"name": "Tanja Muetze",
"role": "Author Response",
"response": "Thank you very much for your thoughtful review. Below we have addressed each of the points raised. The application searches for first-neighbour interactions of molecules in the list presented to it. It did not appear to search for interaction between members of the list, which should not affect the contextual nodes selection much, but will alter the degree-based hubs. This should be commented on, or the documentation made clearer if we are incorrect with this observation. To bypass this problem and make the user more aware of this limitation, the tool should be able to provide more control over how the network is constructed, for example providing the option to exclude first neighbours. This observation is incorrect. CHAT does include interactions between members of the uploaded list. Perhaps what you mean is that interactions between the first neighbors of the uploaded list of genes/proteins are not considered? These are actually also considered in CHAT’s calculations – they are just excluded from the visualization to avoid unhelpful \"hairball\" visualizations. We have now clarified this point in the paper. We are not sure why a user would want to exclude first neighbors – this information is needed in CHAT to identify the contextual hubs. Can this application be made to work with an existing network? We agree that this would be a very nice feature, unfortunately it is actually very difficult to do with the current design of CHAT. CHAT identifies hub nodes that interact with more \"contextual\" nodes than statistically expected using a hypergeometric test. This test is reliant on calculating, N, the number of genes with at least one interaction in the database queried to estimate the background expectation. This parameter would not be known or easily estimated in a user-supplied network as CHAT wouldn't know what database or the data in the database when the network was constructed by the user. One of the nice features of CHAT is that the network is constructed using the latest data available via a PSICQUIC query. The text is entirely gene-centric and may leave an inexperienced use under the impression is is only usable for gene-expression data whereas it is equally useful for the analysis of proteomic data and works with UniProtKB identifiers. Whilst I realise this is apparent to anyone who downloads the app, it may well be worth adding a sentence to both the Summary or Introduction of this paper, and also the description in the App store just to make this very clear to naive users. Good point. We have now edited the text to clarify that protein as well as gene ids can be used to construct the network in CHAT. It may also be worth adding the reference to the 2013 PSICQUIC paper as well as the original as I personally find it more informative and again, may be helpful to the inexperienced user. We have now added the suggested reference to the paper. Again, we thank you for comments and suggestions."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1745
|
https://f1000research.com/articles/5-2104/v1
|
30 Aug 16
|
{
"type": "Study Protocol",
"title": "Applying machine learning to automated segmentation of head and neck tumour volumes and organs at risk on radiotherapy planning CT and MRI scans",
"authors": [
"Carlton Chu",
"Jeffrey De Fauw",
"Nenad Tomasev",
"Bernardino Romera Paredes",
"Cían Hughes",
"Joseph Ledsam",
"Trevor Back",
"Hugh Montgomery",
"Geraint Rees",
"Rosalind Raine",
"Kevin Sullivan",
"Syed Moinuddin",
"Derek D'Souza",
"Olaf Ronneberger",
"Ruheena Mendes",
"Julien Cornebise",
"Carlton Chu",
"Jeffrey De Fauw",
"Nenad Tomasev",
"Bernardino Romera Paredes",
"Cían Hughes",
"Trevor Back",
"Hugh Montgomery",
"Geraint Rees",
"Rosalind Raine",
"Kevin Sullivan",
"Syed Moinuddin",
"Derek D'Souza",
"Olaf Ronneberger",
"Ruheena Mendes",
"Julien Cornebise"
],
"abstract": "Radiotherapy is one of the main ways head and neck cancers are treated; radiation is used to kill cancerous cells and prevent their recurrence.\n\nComplex treatment planning is required to ensure that enough radiation is given to the tumour, and little to other sensitive structures (known as organs at risk) such as the eyes and nerves which might otherwise be damaged. This is especially difficult in the head and neck, where multiple at-risk structures often lie in extremely close proximity to the tumour. It can take radiotherapy experts four hours or more to pick out the important areas on planning scans (known as segmentation).\n\nThis research will focus on applying machine learning algorithms to automatic segmentation of head and neck planning computed tomography (CT) and magnetic resonance imaging (MRI) scans at University College London Hospital NHS Foundation Trust patients. Through analysis of the images used in radiotherapy DeepMind Health will investigate improvements in efficiency of cancer treatment pathways.",
"keywords": [
"Radiotherapy",
"Segmentation",
"Head and Neck",
"Oncology",
"machine learning",
"artificial intelligence"
],
"content": "Background\n\nCancers of the head and neck account for 2% of all cancers, and worldwide they account for around 300,000 deaths each year. Between 1990 and 2006 the incidence of different head and neck cancers has altered dramatically. The incidence of oral cavity cancer has risen by more than 30% to 3.02 per 100,000 and oropharyngeal cancer incidence has more than doubled with a significant change in causation (human papilloma virus rather than smoking or alcohol) and at-risk subpopulation (younger rather than older patients) (NCIN, 2012; OCIU, 2010; Parkin, 2010).\n\nMost such cancers are treated using radiotherapy (CRUK, 2016). Planning such treatment involves delineating the tumour to be irradiated, and also the structures (organs at risk, OAR) to which administration of radiation should be minimised. The complex anatomy of the head and neck, where multiple OARs often lie in extremely close proximity to the tumour, make this process (known as ‘segmentation’) difficult: it can take radiotherapy experts four hours or more to do this (Harari et al., 2010). Furthermore as tumour and body shape change over a course of treatment (which can last weeks), it can be necessary to repeat the segmentation analysis at both fiscal and temporal cost; longer times between cancer diagnosis and treatment increase mortality and worsen outcomes (Chen et al., 2008; Mikeljevic et al., 2004).\n\nAdvances in machine learning have allowed the creation of sophisticated image recognition tools which might perform this process faster and at least as accurately. Machine learning has undergone a revolutionary transformation with the resounding success of so-called Deep Learning algorithms introduced by (Hinton et al., 2006) and demonstrated at scale by others (Krizhevsky et al., 2012). Those algorithms combine artificial neural networks, well-known since their introduction in the mid-50s (Rosenblatt, 1959), with advances in computational power and algorithmics which have enabled remarkable success in handling the deluge of high-dimensional “big data” (Krizhevsky et al., 2012). This resulted in the development and rapid deployment of automatic feature extraction, i.e. combining parts of the data into meaningful elementary units for higher-level processing, that would have previously required the painstaking trial-and-error process of manual design by a human.\n\nSuch processes might be readily applied to automated segmentation for radiotherapy planning in the treatment of head and neck tumours. The prevalence of head and neck cancers, and the complexity of radiotherapy planning, make them ideal targets for such an automated computer-based approach. An automated segmentation system would allow planning to start immediately after a patient is scanned. Such a service could increase speed to treatment. It could also help reduce variation in radiotherapy outcomes between centres (Peters et al., 2010) by standardising planning prior to dose simulation while increasing the efficiency of patient workflow. Although many techniques have been proposed for automatic segmentation in radiotherapy (Daisne & Blumhofer, 2013; Rohlfing et al., 2005), none have shown sufficiently good performance for routine use in clinical care.\n\nIn order for machine learning algorithms to reach expert levels at image segmentation, they must first learn from existing data. This study aims to achieve this with a dataset of expertly labelled images from previously-treated patients at University College London Hospital (UCLH) NHS Foundation Trust (London, UK), with the ultimate objective of improving outcomes for patients with head and neck cancers.\n\n\nAims and objectives\n\n1.1\n\nTo investigate the feasibility of developing computer algorithms that can identify important anatomical structures and the cancers themselves in head and neck cancer planning scans to help target radiotherapy treatment.\n\nShould the primary objective be accomplished, we intend to validate performance using retrospective data through:\n\n2.1\n\nAssessment of quality of automated segmentation using retrospective planning CT images. Expert radiation oncologists, blind to image source, will assess both automatically and manually segmented scans, and determine whether segmentation met a standard for clinical use.\n\n\nStudy design\n\nThis is a retrospective, non-interventional study. Analyses performed in the study will be on fully anonymised medical images (computed tomography (CT) and magnetic resonance imaging (MRI) scans, labelled with manual segmentation, dose threshold and cancer type).\n\nThe protocol follows similar procedures to De Fauw et al. (2016).\n\nPatients who received radiotherapy treatment for head and neck cancers at UCLH NHS Foundation Trust between 01/01/2008 and 20/03/2016 will be eligible for inclusion in this study.\n\nData from patients who have previously manually requested that that their data should not be shared, even for research purposes in anonymised form, and have informed the UCLH NHS Foundation Trust of this, will be ineligible and removed by UCLH NHS Foundation Trust staff before research begins.\n\nApproximately 700 retrospective patient cases will be part of this study.\n\nMost recent machine learning algorithms benefit from large datasets on which to train (tens to hundreds of thousands of data instances (Silver et al., 2016)). Across all machine learning applications the predictive power (as percentage of data instances correctly classified) of the algorithm depends on the size and quality of the dataset.\n\nThe sample size is informed by the existing literature (Mnih, 2015; Silver et al., 2016) and by DeepMinds previous work in the field of machine learning. We believe that the research goals are possible despite the relatively small number of scans, as compared to other research projects because of low variation between the different biological images and by limiting the scope of the research to segmentation rather than diagnosis.\n\nFor all patients meeting inclusion/exclusion criteria the following electronic health record data will be required to complete this project successfully:\n\n(1) CT scan(s) taken during the course of radiotherapy planning and treatment\n\n(2) MRI scan(s) taken during the course of radiotherapy planning and treatment\n\n(3) CT labelling information outlining anatomical and tumour volumes, with associated radiotherapy dose thresholds\n\n(4) Information on what type of tumour is present in each image\n\n(5) Patient gender and age group (to the nearest 5 years)\n\nThe anonymisation procedures adopted will remove any information not specified to further avoid transfer of patient identifiable information. All anonymisation will be formally verified by UCLH NHS Foundation Trust staff before transfer.\n\nIn order to develop the algorithms, DeepMind will work with the labelled medical image files to apply machine learning and AI techniques including but not limited to: supervised and semi-supervised convolutional neural networks, recurrent neural networks, unsupervised clustering, reinforcement learning (Murphy, 2012).\n\nDescriptive statistics (such as the Dice similarity coefficient, average surface distance and maximal surface distance) will be used to compare the quality of algorithm segmentation against the expert reference segmentation during algorithm training.\n\nIn order to assess the accuracy of the final model segmentation a retrospective test subset of radiotherapy planning images will also have both manual and automatic segmentations corrected by expert clinicians who have not seen the images before and who are blinded to how the segmentation was produced. The same statistical methods described above will be used to compare the ground truth manual segmentation, automatic segmentation and clinically corrected manual and automatic segmentations.\n\n\nData protection\n\nThis study requires existing retrospective data only; no prospective data are needed nor will be collected from patients, hospitals or healthcare workers. No direct patient contact will occur and necessary data will be anonymised from this source dataset.\n\nAnonymisation of all image files and clinical information is performed and validated at UCLH NHS Foundation Trust before transfer. No patient identifiers will be transferred to DeepMind. In addition the data will be protected to HSCIC Information Governance standards and access is strictly controlled to prevent any attempt to re-identify the data.\n\nDeepMind Health has developed and established a state-of-the-art secure patient information handling service utilising Common Criteria EAL4 compliant firewalls and on-disk encryption (using Advanced Encryption Standard with a 256-bit key) of all research data, all housed within an ISO 27001 compliant data centre. After anonymisation data will be transferred to our London, UK data centre. This data handling facility conforms to NHS HSCIC Information Governance Statement of Compliance Toolkit (formally assessed at level 3).\n\nAccess will be granted by the custodian of data and no other members of the team. Only those working directly on the data in a research capacity will have access.\n\nThe data sharing agreement between DeepMind and UCLH NHS Foundation Trust lasts for 5 years. After this period the agreement will be reviewed should future work seek to build on this project. After the data sharing agreement expires all data used in the study will be destroyed. No modification will be made based on the data after destruction.\n\nData destruction will involve the deletion of the encryption/decryption keys for all project volumes, and 8-pass random data write to all physical disks within the DeepMind Health data infrastructure. A certificate of destruction will be provided to the Trust.\n\nThe algorithms developed during the study will not be destroyed. DeepMind Health knows of no way to recreate the patient images transferred from the algorithms developed. No patient identifiable data will be included in the algorithms.\n\n\nEthical considerations\n\nThe research on the dataset received formal Research Ethics Committee approval on 6th April 2016 (REC reference 16/SC/0189).\n\nNo patient will be approached directly. Only anonymised retrospective data collected as part of routine clinical care are included. In such cases the ICO code of practice states that explicit consent is not generally required (ICO, 2012).\n\nThe project is non-interventional and does not involve any direct patient contact. All patient data is historical and all patients have completed their radiotherapy treatment prior to data transfer.\n\nThe study will be monitored both internally and externally. Internally DeepMind managers (TB, JL) will oversee and monitor progress on a day-to-day basis, ensuring the protocol is adhered to and no compliance issues arise.\n\nClinical and methodological experts (RM, DD, KS, SAM, GR, RR) are working with DeepMind to further oversee the ethical, clinical and methodological considerations of the project and will advise on at least a weekly basis to ensure no deviation from the described protocol.\n\nExternally the information governance team at the UCLH NHS Foundation Trust will be consulted before commencing data collection.\n\nDeepMind has access to the required data to support the research aims of this study. To ensure compliance with the common law principle of data confidentiality, DeepMind will only receive anonymised data from UCLH NHS Foundation Trust. DeepMind works with the Trust to ensure accuracy and clarity in the data to allow useful and consistent interpretation at all times.\n\nThe results will be disseminated through normal academic channels, initially focusing on conference proceedings and the indexed peer reviewed literature relevant to the fields of machine learning, artificial intelligence, radiotherapy and clinical research. DeepMind will engage in patient and public involvement groups during the research study.\n\n\nConclusion\n\nWe propose an exploratory study covers an initial testing of machine learning algorithms for automatic segmentation of head and neck planning CT and MRI scans. The results will be assessed against expert segmentation.",
"appendix": "Author contributions\n\n\n\nAll authors contributed to study design and methodology. CC, CH, JD, NT, BP, OR and JC contributed to machine learning approaches. RM, KS, DD and SAM contributed expertise in oncology, radiation physics and radiography. TB and JL contributed to project steering and information governance. GR, RR and HM contributed to methodological oversight.\n\n\nCompeting interests\n\n\n\nUniversity College London Hospital NHS Foundation Trust administration time spent on this work will be paid to the Trust by DeepMind.\n\nThe Chief Investigator (CC) and some co-investigators (TB, JC, CH, OR, JD, NT, BP) are full time employees of DeepMind. GR, RR, HM and JL are paid consultants of DeepMind. The company is also funding the research.\n\n\nGrant information\n\nDeepMind is the sole funder. No grants were involved in supporting this research.\n\n\nAcknowledgments\n\nWill Kay is the DeepMind data custodian and is responsible for protection and security of the dataset described in this protocol.\n\n\nReferences\n\nCancer Research UK: Oral Cancer Statistics. Cancer Research UK. 2016; (last accessed 16/05/2016). Reference Source\n\nChen Z, King W, Pearcey R, et al.: The relationship between waiting time for radiotherapy and clinical outcomes: A systematic review of the literature. Radiother Oncol. 2008; 87(1): 3–16. PubMed Abstract | Publisher Full Text\n\nDaisne JF, Blumhofer A: Atlas-Based automatic segmentation of head and neck organs at risk and nodal target volumes: a clinical validation. Radiat Oncol. 2013; 8: 154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Fauw J, Keane P, Tomasev N, et al.: Automated analysis of retinal imaging using machine learning techniques for computer vision [version 1; referees: 1 approved]. F1000Research. 2016; 5: 1573. Publisher Full Text\n\nHarari PM, Song S, Tomé WA: Emphasizing conformal avoidance versus target definition for IMRT planning in head-and-neck cancer. Int J Radiat Oncol Biol Phys. 2010; 77(3): 950–958. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHinton GE, Osindero S, Teh YW: A Fast Learning Algorithm for Deep Belief Nets. Neural Comput. 2006; 18(7): 1527–54. PubMed Abstract | Publisher Full Text\n\nInformation Commissioner's Office: Anonymisation: managing data protection risk code of practice. 2012; (last accessed 03/06/2016). Reference Source\n\nKrizhevsky A, Sutskever I, Hinton GE: ImageNet Classification with Deep Convolutional Neural Networks. In NIPS. 2012; 1: 4. Reference Source\n\nMikeljevic JS, Haward R, Johnston C, et al.: Trends in postoperative radiotherapy delay and the effect on survival in breast cancer patients treated with conservation surgery. Br J Cancer. 2004; 90(7): 1343–1348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMnih VM, Hinton GE: Learning to Detect Roads in High-Resolution Aerial Images. European Conference on Computer Vision. 2010; 6316: 210–223. Publisher Full Text\n\nNCIN: Potentially HPV-related head and neck cancers. NCIN Data Briefing. 2012. Reference Source\n\nOxford Cancer Intelligence Unit: Profile of Head and Neck Cancers in England: Incidence, Mortality and Survival. 2010. Reference Source\n\nParkin DM, Boyd L, Walker LC: 16. The fraction of cancer attributable to lifestyle and environmental factors in the UK in 2010. Br J Cancer. 2011; 105(Suppl 2): S77–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeters LJ, O’Sullivan B, Giralt J, et al.: Critical impact of radiotherapy protocol compliance and quality in the treatment of advanced head and neck cancer: results from TROG 02.02. J Clin Oncol. 2010; 28(18): 2996–3001. PubMed Abstract | Publisher Full Text\n\nRohlfing T, Brandt R, Menzel R, et al.: Quo vadis, atlas-based segmentation? In The Handbook of Biomedical Image Analysis: Segmentation and Registration Models. Edited by Suri J, Wilson DL, Laxminarayan S. New York, NY: Kluwer Academic/Plenum Publishers. 2005; 435–486. Publisher Full Text\n\nRosenblatt F: The perceptron: a probabilistic model for information storage and organization in the brain. Psychol Rev. 1958; 65(6): 386–408. PubMed Abstract | Publisher Full Text\n\nSilver D, Huang A, Maddison CJ, et al.: Mastering the game of Go with deep neural networks and tree search. Nature. 2016; 529(7587): 484–489. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "17312",
"date": "31 Oct 2016",
"name": "Yuan Feng",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study protocol presented the procedure to acquire and analyse medical image data for automatic segmentation. The protocol is clear besides the following points need to be clarified:\nThe authors presented the algorithm used for segmentation very briefly. This part should be elaborated to inform the details of each algorithm used. Also, citation of Murphy 2012 was not found in the reference.\n\nThe statistical analysis was not well presented. The evaluation and statistical methods should be described in detail.\n\nThere are many different segmentation studies in the literature, some specifically aimed for radiotherapy should be cited.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2104
|
https://f1000research.com/articles/5-2082/v1
|
26 Aug 16
|
{
"type": "Research Article",
"title": "A Bayesian spatial random effects model characterisation of tumour heterogeneity implemented using Markov chain Monte Carlo (MCMC) simulation",
"authors": [
"Martin D. King",
"Matthew Grech-Sollars",
"Matthew Grech-Sollars"
],
"abstract": "The focus of this study is the development of a statistical modelling procedure for characterising intra-tumour heterogeneity, motivated by recent clinical literature indicating that a variety of tumours exhibit a considerable degree of genetic spatial variability. A formal spatial statistical model has been developed and used to characterise the structural heterogeneity of a number of supratentorial primitive neuroectodermal tumours (PNETs), based on diffusion-weighted magnetic resonance imaging. Particular attention is paid to the spatial dependence of diffusion close to the tumour boundary, in order to determine whether the data provide statistical evidence to support the proposition that water diffusivity in the boundary region of some tumours exhibits a deterministic dependence on distance from the boundary, in excess of an underlying random 2D spatial heterogeneity in diffusion. Tumour spatial heterogeneity measures were derived from the diffusion parameter estimates obtained using a Bayesian spatial random effects model. The analyses were implemented using Markov chain Monte Carlo (MCMC) simulation. Posterior predictive simulation was used to assess the adequacy of the statistical model. The main observations are that the previously reported relationship between diffusion and boundary proximity remains observable and achieves statistical significance after adjusting for an underlying random 2D spatial heterogeneity in the diffusion model parameters. A comparison of the magnitude of the boundary-distance effect with the underlying random 2D boundary heterogeneity suggests that both are important sources of variation in the vicinity of the boundary. No consistent pattern emerges from a comparison of the boundary and core spatial heterogeneity, with no indication of a consistently greater level of heterogeneity in one region compared with the other. The results raise the possibility that DWI might provide a surrogate marker of intra-tumour genetic regional heterogeneity, which would provide a powerful tool with applications in both patient management and in cancer research.",
"keywords": [
"Tumour heterogeneity",
"diffusion",
"random-effects",
"MCMC",
"Markov chain Monte Carlo",
"Bayesian",
"spatial statistical modelling"
],
"content": "Introduction\n\nNumerous investigations have demonstrated a surprising level of intra-tumour heterogeneity in a variety of cancers1–3. In particular, a picture is emerging in which intra-tumour genetic regional differences can be as great as those observed between cases. It has become widely accepted that spatial heterogeneity underlies the tumour evolutionary process itself. Thus, tumour growth is conceived as a Darwinian process in which spatially heterogeneous mutations occur. The implications are enormous, and intra-tumour heterogeneity poses challenges and questions for those searching for effective treatments. For example, it has been suggested that drug resistance is an inevitable consequence of intra-tumour genetic diversity, and that the presence of many different genomes increases the probability that a particular population of cells develop resistance. It is also suggested that a given drug might kill a majority of tumour cells, leaving those that are resistant to become dominant in a Darwinian-like selection processes. Thus, according to this proposition, selection is driven by the treatment itself. Furthermore, clinical decision making and patient management based on a standard biopsy must be questionable and the very notion of personalised medicine might be a greater challenge than initially conceived. The realisation that treatment can drive the evolutionary process might indicate a need to revise current treatment strategies4.\n\nMany papers have appeared in the biomedical imaging literature dealing with tumour spatial heterogeneity and proposing a variety of methods for characterising the resulting distributions. Quantile estimates are commonly used for this purpose, but a variety of other methods have been adopted including, for example, measures based on the departure of the observed data from a simple idealised spatial structure5 and functional principal components analysis6. A key feature of the majority of these analyses is their two-stage nature in which voxel-by-voxel parameter estimates are derived from the image data, followed by a second-stage analysis of the resulting parameter distribution. The present paper outlines a single-stage approach to characterising tumour heterogeneity in DWI images based on so-called random effects modelling. A key feature of the method is that a formal spatial model is included in the statistical procedure used to extract the diffusion parameter estimates from the signal intensity data. In fact, given the paucity of the DWI data used in the present analyses, the spatial random effects treatment is an indispensable component of the estimation method.\n\nA key tenet underlying the application of diffusion MRI (dMRI) to cancer patient management is the notion that the apparent diffusion coefficient (ADC) is a surrogate marker of altered cellularity. Despite numerous investigations into the relationship between cellularity and MRI surrogate markers7–14 several issues remain to be resolved. One might question whether positive tests of association, including correlation coefficient tests, are sufficient to justify the surrogate-marker claim. A relatively weak relationship can be sufficient to yield a statistically significant test result. It might be argued that the requirements of a biomarker/surrogate marker should meet the criteria of a surrogate endpoint in clinical trials (see, for example, 15). Secondly, some of the published evidence is based on statistical models that might be judged inadmissible. Furthermore, some researchers use p-values as the sole supporting evidence of biomarker validity. Although p-values may provide a measure of the strength of evidence (recognising that this contravenes the rules governing the frequentist approach to statistical inference), a p-value does not provide a measure of the strength of the effect/association under consideration (see, for example, 16). Among the complications is the p-value dependence on sample size. Moreover, the p-value provided by an association test does not address the key issues of specificity and sensitivity.\n\nWorking under the assumption that the ADC is a valid surrogate marker of cellularity, researchers have focussed on a variety of tumour ADC measures, including various global spatial heterogeneity and dispersion/distribution indicators (see, for example 9 or 17). Some have examined the spatial dependence of the ADC in the vicinity of the tumour boundary (see, for example, 14, or 17) while others have focussed their efforts on the elucidation of the underlying causes of the diffusion changes11,18, or have adopted an entirely empirical approach and examined the relationship between survival and one or other DWI measure19–22. The latter studies aim to address the central clinical issue and provide a direct answer to the fundamental question regarding the prognostic/diagnostic value of dMRI. This literature provided the motivation for the present statistical modelling work. Although the mechanistic basis of the ADC changes that occur during tumour development may remain elusive, this does not preclude the possibility that dMRI might have the potential to provide a useful prognostic indicator. In keeping with this empirical approach, the present study was undertaken to develop a formal statistical modelling procedure for tumour DWI spatial heterogeneity estimation. The main focus is robust voxel-specific ADC estimation and an examination of the ADC dependence on distance from the tumour boundary. This is prompted by a number of reports indicating that some tumours exhibit a boundary-distance dependence in ADC, the magnitude of which might carry prognostic information14,17,22. In particular, we examine the magnitude and significance of the boundary-distance effect that remains after simultaneously adjusting for an underlying random 2D heterogeneity in ADC. In addition, heterogeneity in the boundary region is compared with that in the core. As stated above, this is based on the premise that empirical measures might provide useful prognostic information in the absence of an understanding of the mechanistic basis of the spatial variation and temporal changes in diffusion that have been observed to occur in a variety of tumours.\n\nThe tumour heterogeneity analysis outlined in this paper is based on a Bayesian spatial random effects (random coefficients) analysis of paediatric DWI data. Key to this approach is the provision of robust estimates of the voxel-specific diffusion parameters, as required to obtain reliable measures of spatial heterogeneity. The random coefficient estimates are expected to be more robust than the voxel-specific parameter estimates provided by an independent-voxels (separate voxel-by-voxel) analysis. In common with all Bayesian analyses, prior distributions are a central part of the statistical model. These provide a formal mechanism for incorporating existing information and model assumptions. In the present study Markov random field prior distributions were adopted for a number of parameters, including the voxel-specific anisotropic diffusion coefficients, as outlined below and in the Methods section. The reason why Bayesian random effect models have the potential to provide improved parameter estimates, relative to those given by an independent-voxels analysis, is because they make good use of the available data through so-called information borrowing. In the present context, the signal behaviour in adjacent voxels influences the parameter estimates obtained for each voxel under consideration. Formal borrowing of information across a region and the resulting improvement in estimation is mediated via the prior distributions that are assigned to each model parameter. Restated, the distributional priors underlie the information borrowing that is key to random effects modelling. (The distributional assumptions required to construct an hierarchical random effects model are priors, by definition, regardless of the analytical framework, be it frequentist or Bayesian.) The resulting estimators are referred to as shrinkage estimators, the aim being to provide some shrinkage towards the typical behaviour, and thus achieve some level of smoothing. Specifically, shrinkage refers to the condition where more extreme estimates are pulled towards more typical values, as determined by the distribution characteristics (spatial correlation structure in the present application) of the ensemble of units (voxels in the present study) under consideration. More robust estimates of the underlying and unknown parameters are thus obtained, improving on those that might be derived from an independent units (isolated voxel-by-voxel) analysis. Necessarily, one accepts a trade-off between bias and improved variance. Nevertheless, in any situation in which there is a true, non-negligible underlying variation between the units under consideration, combined with a non-negligible measurement error, neither completely-pooled estimation (averaging over the ROI) nor the estimates obtained through a set of separate analyses are uncompromised. This issue is discussed in Sections 5.4 and 5.5 of the textbook by Gelman et al.23, where they use a simple dataset to compare the results given by a random effects treatment with the completely pooled result and an independent units analysis. They make a convincing case for random effects modelling.\n\nAmong the research disciplines in which Bayesian spatial random effects modelling is especially prominent is epidemiological disease mapping. Disease mapping and the present tumour heterogeneity study share a similar objective, namely the extraction of an underlying spatial structure, given data that are typically corrupted by noise. Sparseness/rarity of the observed events is a problem in some disease mapping applications, which is not dissimilar to the sparse data problem that arises in the present study. Markov random field priors are common among those adopted in the disease mapping literature24–26. Note S1 provides a brief introduction to Markov processes and Markov random field models which, together with references given in the Methods section, serves as an entry point to the literature. The analyses outlined in this paper were performed using a so-called conditional autoregressive (CAR) form of Markov random field prior, as explained in Note S1.\n\nGiven a Bayesian hierarchical random effects model, some procedure is required for computing the posterior probability distribution. This invariably involves complicated high-dimensional integrals that have no analytical solution. MCMC is often adopted as a method that circumvents the analytical intractability of this kind of problem23,27,28. The Gibbs sampler is among the most widely used MCMC algorithms; it is based on an iterative sampling of a set of conditional distributions. The CAR prior referred to in the preceding paragraph (a conditional distribution by definition), thus fits naturally into the Gibbs sampler algorithm, and the resulting computational efficiency is among the appealing features of adopting this prior when performing a Bayesian spatial analysis using the Gibbs sampler. Computer software for performing Bayesian spatial data analyses is readily available, including Gibbs sampler implementations. The MCMC analyses outlined in this paper were performed using WinBUGS/GeoBUGS29,30.\n\nIn summary, the purpose of the present study was to develop a model for tumour ADC spatial heterogeneity. This was motivated by current biomedical research indicating that tumour heterogeneity has important implications in the search for improved cancer treatment strategies and for the investigation of tumour pathophysiology1–3. We have adopted a spatial random effects modelling approach to characterising heterogeneity, implemented using MCMC. Despite the merits of performing an MCMC analysis, the method is not infallible. Reliable statistics depend on achieving convergence to a stationary distribution. Convergence assessment is, therefore, an essential part of any MCMC analysis. In order to guard against misleading heterogeneity measures, we have paid reasonable attention to the convergence issue, in addition to addressing the question of model adequacy. The latter was achieved by using posterior predictive simulation to examine key features of the spatial statistical modelling results, as described in the Methods section.\n\n\nMethods\n\nThe data used in this study are the same as those used in a previous study22 for which ethical approval was granted by the National Research Ethics Service Committee London – Bloomsbury. Therefore, no additional approval was required. Patient details and related information are given in 22. The subset of patients selected for the present work were imaged using a Siemens Magnetom Symphony scanner, capable of generating magnetic field gradients of amplitude up to 30 mT m–1. DWI data were acquired using a diffusion-sensitized single-shot echo planar imaging sequence (acquisition matrix 128 × 128, image matrix 256 × 256, field-of-view 230 × 230 mm, twenty 5mm slices separated by 2.5mm, TR 3600 ms, TE 107 ms). In addition to a single b0 image, 6 diffusion-weighted images were acquired with b-values 500 and 1000 s mm–2 for each of 3 orthogonal directions. The total imaging/sequence time was 56s.\n\nThe data were not formally blinded because the investigation does not take the form of a clinical trial. Instead, the purpose of the study was to develop a statistical model, with parameter estimation as the objective, focussing on tumour heterogeneity measures. Thus the goal is parameter estimation as distinct from hypothesis testing. That said, the model development and data analyses were performed by MDK, using image signal intensity data provided in NIFTI format, with all sources of identification removed. The FSL utility tools (Version 4.1.5) fslview, fslslice and fsl2ascii (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Fslutils) were used to identify the location of each tumour, and to extract the DWI signal intensity data from the tumour region, with conversion to ASCII format.\n\nRegion-of-interest selection was based on an inspection of ADC maps. A core ROI of 15-by-15 voxels was placed at a position well removed from the tumour boundary, as shown for one case in Figure 1. Tumour boundary ROIs were selected in that portion of the tissue where the boundary is easily identified due to the presence of adjacent oedematous voxels. A subject-specific ADC threshold was chosen and used to define tumour versus extra-tumour voxels. These thresholds were based on an inspection of the ADC values in oedematous voxels, together with a visual examination of the segmentation obtained with alternative values. ADC profile plots were used to select the boundary ROI width, as judged by the number of voxels over which the boundary distance effect vanishes, and the ADC becomes indistinguishable from typical core values. Boundary distance was equated to the minimum of the row and column distances. (No marked differences occurred when the analyses were performed using minimum Euclidean distances.) Figure 1 includes a magnified portion of the ADC image to show the boundary ROI in greater detail. It also shows part of the corresponding array of ADC values. In this case a threshold of 117 was used to distinguish tumour versus oedematous voxels. This study is of an exploratory nature, and tumour boundary identification and ROI selection are among the issues that require refinement, given the fact that the tumour boundary is never well defined in its entirely. We return to the ROI selection and automation problem in the Discussion.\n\nThe left panel shows an ADC image with the core ROI (square region) and boundary ROI (irregular region) superimposed. The boundary ROI is shown with increased magnification in the upper right panel. The array of ADC values given in the lower right panel are taken from the subregion indicated by the white box. The lines superimposed on the ADC array correspond to the ROI boundary.\n\nThe main purpose of the modelling study outlined in this paper is to determine whether DWI data obtained from PNET (primitive neuroectodermal tumour) cases provide statistical evidence to support the proposition that water diffusivity in the boundary region of some tumours exhibits a dependence on distance from the boundary, in excess of an underlying 2D spatial heterogeneity in diffusion. Tumour ADC dependence on boundary-distance has been reported in previous publications, and the principal objective is to determine whether this effect can be demonstrated given a model that includes additional 2D spatial variation. The distinction is between a monotonic change in diffusion in a direction approximately normal to the boundary, compared with an underlying and general 2D spatial dependence, with no distinct orientation relative to the tumour boundary. Specifically, an inverse tumour boundary-distance term is included in the diffusion model, together with additional random effect diffusion terms to capture the underlying 2D spatial heterogeneity (spatial correlation structure). Separate random effect terms are assigned to each of the three (read, phase and slice) diffusion coefficients. Interest focusses on the magnitude and statistical significance of the inverse-distance coefficient, in order to determine whether a formal spatial model provides robust evidence for a boundary effect. The statistical model allows the boundary voxels to differ from core voxels in their spatial correlation/heterogeneity, thus facilitating a comparison of the two regions. Furthermore, the model deals with the complication that arises from non-monoexponential signal-intensity dependence on gradient amplitude, combined with an additional limitation arising from our using existing DWI data acquired with a standard clinical image acquisition protocol. The main limitation of the clinical data is a lack of replication. Thus each DWI dataset consists of 7 images only, namely, a single b0 image plus single acquisitions with gradient sensitisation in the x, y and z directions and b-values of 500 and 1000 s mm–2. The lack of b0 data replication coupled with the non-monoexponential decay is the main challenge because a standard analysis cannot yield estimates of the true b0 signal intensities separated from the accompanying noise. An additional limitation of the clinical data is the restriction to two non-zero b-value observations in each direction, again with no replication. The required diffusion parameter estimates can be obtained, however, under the spatial model outlined below, despite the departure from exponential dependence on magnetic field gradient amplitude and lack of replication. In essence the model assumes an exponential dependence in a b-value range that includes the b500 to b1000 observations, and captures the departure of the b0 signal intensity from the exponential curve using an offset parameter (δi in equation 3). The latter is incorporated in the form of a random effect term, and it is among those that are assigned a spatial CAR prior distribution (the CAR prior is outlined in Note S1). The essential features of the measurement model are illustrated in Figure 2. To re-iterate, an offset term captures the low b-value signal decay and yields a b0 signal intensity estimate, separated from the noise contribution, despite the lack of b0 replication. The b0 signal intensity estimates are subject to two constraints, namely the underlying spatial distribution referred to above, together with a measurement error distribution with variance equal to the error variance of the finite b-value observations. This approach is preferable to treating the b0 observations as error free. The resulting voxel-specific signal intensity and diffusion parameter estimates were used to calculate a summary parameter, ADC0.5 (given the symbol d′ in the following equations), based on the half-maximum-intensity b-value estimates, thus circumventing the non-monoexponential decay problem. The model details are as follows.\n\nSignal attenuation from a b-value of 380 s mm–2 up to 1000 s mm–2, and beyond, is assumed to be exponential, with a voxel-specific diffusivity characterised by ADCs (slow diffusion ADC component). The model allows for a departure from mono-exponential behaviour, which is assumed to occur below a b-value of 380 s mm–2. (The need to assume that the mono-exponential behaviour extends down to 380 s mm–2 arises because some of the b^0.5 estimates (definition given below), which are used to calculate the ADC0.5 estimates, are less than 500 s mm–2, the lowest falling at approximately 380 s mm–2. In contrast, the mono-exponential assumption underlying the ADCs calculations is restricted to the b-value range 500 s mm–2 to 1000 s mm–2.) An offset term, δ, captures the additional signal attenuation that occurs over the b-value range 0 to 380 s mm–2, thus accounting for the expected departure from mono-exponential behaviour at low b-value. δ is the difference between the true b0 signal intensity and the intercept, S, given by the mono-exponential expression for signal intensity. The dashed curve shows this exponential behaviour extrapolated to zero b-value. As outlined in the Statistical model subsection of the Methods section, ADC0.5 is calculated using b^0.5=log(2S/μ(0,k↑))/d¯, where b^0.5 is the b-value at the half-maximum signal intensity, noting that the subscript labels used in the main text (indicating that each of these parameters is voxel specific) have been dropped for the sake of simplicity. Similarly, the associated random effect terms are not included in the schematic.\n\nThe measurement model takes the form\n\nIn summary, the model includes a spatial random effects term (θi1) to cater for between-voxel variation in the b0 intercept of the slow-diffusion component (i.e., Si in equation 2), specified as having a conditional variance common to both boundary and core voxels, a spatial random effect term (θi2) to allow for between-voxel variation in the fast-diffusion component, i.e., the departure from mono-exponential decay, again with a conditional variance common to the two regions, and spatial random-effect contributions (θik, k = x, y, z) to the magnetic field gradient direction-specific diffusion coefficients, with separate conditional variances assigned to the core and boundary and to each direction. A diffusion summary parameter (d′) was adopted to circumvent the complication that arises due to non-exponential b-value dependence. This is based on the assumption that the direction-specific mono-exponential expression for signal attenuation given in equation 2 applies over the range b380 to b2000. (With the exception of a few outliers with larger values, the voxel-specific b^0.5i estimates, referred to below, lie within this b-value range. It might be noted that a proportion of these estimates lie outside the b500 and b1000 range, which is not ideal. This problem is a reflection of our using retrospective clinical data obtained with an imaging sequence that was designed for routine DWI, as opposed to detailed statistical modelling.) Accordingly, a voxel-specific half-maximum signal-intensity b-value was calculated using b^0.5i=log(2Si/μi(0, k↑))/di, where di = (dix + diy + diz)/3 is the mean diffusivity (the subscript i is the voxel label, not to be confused with the convention where it is used to indicate an isotropic parameter). This is used, in turn, to calculate the summary parameter d′i = log(2)/b^0.5i. For the sake of conciseness and readability the voxel subscript is dropped in the Results and Discussion sections, and the abbreviations ADCs, ADC0.5 and µ0 are used for di, di′ and µi (0, k ↑), respectively, i.e., the slow-diffusion ADC component, the ADC based on an estimate of the half-maximum signal intensity, and the b0 signal intensity estimate. A schematic showing the relationship between the key model parameters and the b-value dependence in signal intensity is shown in Figure 2.\n\nThe summary parameters ADCs and ADC0.5 were calculated using the definitions given above. The various range statistics were generated by evaluating the relevant minimum and maximum voxel-specific values at each MCMC iteration. Thus the resulting statistics include all sources of variation/error, including uncertainty in the identity of the voxels responsible for the extreme values. The boundary-region 2D heterogeneity ADCs statistics listed in Table 2 were derived from the average of the three magnetic field gradient-specific diffusion coefficients, as obtained for each voxel with the boundary effect (β/li) removed (i.e., based on dik = µdkb + θik, i = Nc + 1, . . ., NT; k = x, y, z; compare with equation 8). The third data column in Table 2 lists the boundary distance-effect (β/li, i = Nc + 1, . . ., NT) ranges, as given by the difference between the maximum and minimum voxel-specific values.\n\nBayesian spatial modelling and its implementation using MCMC is well documented (see, for example, 24–26,31). We have previously demonstrated the application of this modelling approach to the crossing-fibre problem that arises in diffusion tensor imaging32 and the sparse data problem that often arises in MR perfusion and diffusion image analyses33,34. Gibbs sampling was performed using the WinBUGS/GeoBUGS package (Version 1.4.3)29,30,35,36, which was downloaded from http://www.mrc-bsu.cam.ac.uk/bugs. Three parallel chains were generated for each of the five analyses (5 cases examined in total), each chain consisting of 5000 samples after thinning. (Thinning is the name given to the common practice whereby a specified proportion of the MCMC output is discarded, the remaining samples being stored for subsequent processing. It has been discussed by a number of analysts, including Carlin and Louis (2009) (Section 3.4.5 in 28). The reason for thinning the MCMC output is to produce a chain with reduced sample autocorrelation. The aim is to reduce both the storage and post-simulation CPU demands of an analysis without suffering much loss in precision. A thinning factor of 40 was used in the majority of the simulations, excepting the posterior predictive analyses, which were performed using a thinning factor of 10. Thus, 1 in 40 (or 1 in 10) samples was stored and used in subsequent calculations). The first of each set of three chains was started at an arbitrary position in parameter space, while the other two were started at over-dispersed positions. A burn-in set of samples was acquired prior to storing each chain of 5000 samples. The burn-in samples were discarded.\n\nAn informal assessment of convergence was performed by inspecting selected overlaid trace plots for visual signs of convergence failure. (MCMC convergence analysis is a topic that has been discussed by many analysts, as summarised in several textbooks, (Gelman et al23, and Carlin and Louis28, for example) and discussion papers37,38. Our approach is based partly on their recommendations.) This visual assessment was followed by a semi-formal analysis (see 39 for a review of the methods) which was performed using three convergence test procedures, namely the Gelman-Rubin shrink factor diagnostic and associated shrink factor plots (which is based on an ANOVA-like assessment of the between-chain and within-chain variances), the Geweke Z-score diagnostic and Z-score plots (based on spectral density variance estimation and a Z-score comparison of chain segments), and the Raftery-Lewis diagnostic procedure (which provides a variety of data, including an estimate of the number of iterations required to obtain a given quantile to some specified accuracy, taking into account the correlation between samples). Particular attention was paid to the accuracy obtained for the key measures of spatial variation. These convergence analyses were performed using the R CODA (Convergence Diagnosis and Output Analysis) package40 (R version 2.15.2, CODA version 0.16-1).\n\nModel assessment was undertaken using posterior predictive simulation (41–43; Chapter 6 in 23, provides a useful introduction.) This is an established procedure which can be used to calculate so-called posterior predictive p-values or Bayesian p-values, which serve to determine whether some aspect of the data is unexpected under the model, indicating potential model inadequacy. The objective is to adopt tests that probe the capacity of the model to capture features in the data that are key to the scientific question underlying the research (see, for example, page 172 in 23). In the present study the resulting Bayesian p-values are used as probabilistic measures of the extent to which the observed signal intensities are more extreme than the posterior predictive data (yrep). In short, the p-values provide a measure of discrepancy. Following Gelman et al.23, page 162, and noting that in the present analysis the test quantity depends only on the data,\n\nIn those instances where this yields a p-value greater than 0.5, the sign of the test is reversed. Thus improbable test results are indicated by Bayesian p-values near zero.\n\nThe preceding mathematical expressions differ from those given in 23, the latter using discrepancy measures that depend on unknown parameters, in addition to the data. In the present study, posterior predictive tests were all performed using signal intensity in isolation as a measure of discrepancy. (Thus T(y, θl), as given on page 163 of 23, becomes T(yi) = yi, where i is the voxel label.) The rationale behind focusing on discrepancies between the observed signal intensities and replicate data generated under the model was as a simple procedure to detect potential outliers, and to determine whether these are associated with extreme ADC estimates, leading to spurious heterogeneity measures, noting that tumour heterogeneity is the focus of the study.\n\n\nResults\n\nThe Results section of this paper is organised as follows. The first subsection reports the main findings of the tumour spatial heterogeneity analysis, which compares the level of heterogeneity observed in the core and boundary region of five tumours. This is followed by two subsidiary sections, the first of which examines the boundary distance effect in greater detail, followed by a subsection dealing with potential anisotropic behaviour in the observed spatial heterogeneity. The final subsection provides a brief summary of the MCMC convergence and simulation accuracy analyses, together with a summary of the posterior predictive simulation results.\n\nThe main motivation for the ADC analysis outlined in this paper is the MRI literature suggesting that the ADC in the boundary region of some tumours exhibits a voxel-specific dependence on distance from the tumour boundary14,17,22. The aim was to determine whether an analysis based on a formal spatial model provides supporting evidence for a boundary effect after adjusting for an underlying random 2D spatial heterogeneity. In this context the statistical model accommodates several parameters of potential interest. These include a term, specific to the boundary ROI, which assumes an inverse dependence of the anisotropic diffusion coefficients on distance from the tumour boundary. This effect is superimposed on an underlying random 2D spatial heterogeneity that is assumed to exist in both the boundary and core regions. The latter general heterogeneity is incorporated in the form of so-called spatial random effect terms. The magnitude of the resulting 2D spatial heterogeneity is allowed to differ between the two regions. Interest focusses on potential subject-specific differences in heterogeneity between the boundary and core regions, together with the magnitude of the boundary distance effect. A note of explanation might assist readers unfamiliar with the Bayesian terminology used in this paper. Each of the model parameters is estimated with error. This uncertainty is captured by the posterior probability distribution. The posterior median provides a point estimate of the true parameter value, while the tail quantiles give an indication of the uncertainty in the estimate. 95% posterior intervals are given for every parameter estimate, including the various range parameters which are used as measures of spatial heterogeneity. In the latter case, the 95% posterior intervals provide a measure of the range of uncertainty in the ADC range estimates.\n\nTable 1 lists the minimum and maximum voxel-specific ADCs estimates, as obtained for the core and boundary regions in five subjects, together with the corresponding range statistics. The main result that emerges from this table is that heterogeneity in the core is not consistently less than that observed in the boundary region, despite the additional boundary distance-effect contribution to the latter. Thus, although heterogeneity in the boundary region ADCs is substantially larger than that observed in the core in 3 of the 5 cases, in one of the remaining cases (Subject 2) the core region exhibits a greater level of heterogeneity.\n\n*The median of the posterior distribution obtained for each of the specified ADCs parameters is listed, together with the 0.025 and 0.975 quantiles, which are given in brackets.\n\nCore and boundary-region minimum and maximum voxel-specific ADCs estimates are listed, together with the corresponding ranges (difference between the maximum and minimum of the voxel-specific estimates). The ADCs range estimates, which are used as measures of ADCs dispersion, are derived from the raw MCMC sample, as opposed to subtracting the listed minimum and maximum values, hence the apparent discrepancy between some of the extreme-value and corresponding range statistics.\n\nTable 2 provides estimates of the magnitude of the random 2D spatial heterogeneity contribution to the boundary-region variation in diffusivity, as captured by the spatial random effects, together with the variation attributable to the deterministic boundary-distance effect, which operates close to the boundary, and the magnitude of the random 2D spatial heterogeneity in the core. Again, these are expressed in the form of ADCs range statistics, as given by the difference between the minimum and maximum voxel-specific ADCs values. The core-region range statistics are included for completeness, although these are identical to those given in Table 1. No consistent pattern emerges from a comparison of the boundary and core spatial random effect ranges (ie the random 2D heterogeneity effect). In particular, there is no indication of a consistently greater level of heterogeneity in one region compared with the other. A comparison of the boundary distance-effect ranges with the corresponding boundary random 2D heterogeneity ranges suggests that both are important sources of spatial variation, although the 2D heterogeneity component dominates in two of the five cases (Subjects 4 and 5).\n\n*The median of the posterior distribution obtained for each ADCs range estimate (10−3 mm2 s−1) is listed, together with the 0.025 and 0.975 quantiles, which are given in brackets.\n\nRange statistics are used as measures of ADCs dispersion in order to determine the relative contribution of the underlying sources of spatial variation. ADCs dispersion in the boundary region is composed of a random 2D spatial heterogeneity contribution and a deterministic boundary-distance effect which operates close to the tumour boundary. In contrast, dispersion in the core is restricted to a random 2D heterogeneity effect, which is permitted to differ in magnitude to the 2D spatial variation observed within the boundary region. Core region ADCs ranges are listed in the first data column, as determined by the difference between the maximum and minimum voxel-specific ADCs values within the selected region. ADCs range estimates attributable to the boundary-region 2D spatial heterogeneity component are listed in the second data column. The third data column lists the contribution of the boundary-distance effect, again expressed as an ADCs range equal to the difference between the voxel-specific maximum and minimum values within the boundary region.\n\nHaving examined the relative magnitude of the ADCs heterogeneity in the boundary-region with that in the core, Table 3 focusses on ADC0.5. ADC0.5 (denoted d′ in the Methods section) is a diffusion summary parameter that captures the signal intensity departure from an exponential dependence on b-value as it approaches zero, in addition to the slow diffusion component. It is based on an estimate of the b-value at the half-maximum signal intensity. (In contrast, ADCs is based on the monoexponential expression for diffusion (slow diffusion component) that is assumed to apply at b-values between 500 s mm–2 and 1000 s mm–2. The ADC0.5 calculation supposes that the departure from mono-exponential dependence occurs below 380 s mm–2. Details of the ADCs and ADC0.5 estimation method are given in the Methods section.) Again, no consistent pattern emerges, noting that in one of the five subjects (Subject 3) the core region appears markedly more heterogeneous than the boundary region.\n\n*The median of the posterior distribution generated for the specified ADC0.5 parameter is listed, together with the 0.025 and 0.975 quantiles, which are given in brackets.\n\nThe core and boundary-region minimum and maximum voxel-specific posterior ADC0.5 (denoted d′ in the Methods section) estimates are given, together with the corresponding ranges. ADC0.5 is derived from the half-maximum signal intensity point on the mono-exponential decay curve as outlined in the Methods section. Spatial variation in the boundary region includes contributions from a random 2D heterogeneity in ADCs (captured by the diffusion coefficient random effect terms), a deterministic boundary-distance effect on diffusion and a boundary distance contribution to heterogeneity in the b0 signal intensity. The range statistics are derived from the raw MCMC sample, as opposed to subtracting the listed maximum and minimum values, hence the apparent discrepancy between some of the extreme value estimates and corresponding range statistics.\n\nThe results given in the preceding section indicate that the boundary-distance effect on diffusion (captured by the term β/l in equation 8) makes an important contribution to the spatial heterogeneity in ADC that is observed in the vicinity of the tumour boundary. In this section the effect is examined in greater detail. In addition to diffusion-coefficient dependence on boundary distance, an inverse distance term has also been added to the expression for the b0 signal intensity in the boundary region (α/l in equation 6), although there is no prior reason for assuming that this term will be important. The regression-parameter posterior median estimates are given in Table 4, together with the corresponding posterior 0.025 and 0.975 quantiles. With the exception of the posterior interval listed for α in Subject 2, the remaining 0.95 posterior intervals all exclude zero. This provides statistical evidence for the presence of a boundary distance effect on the slow diffusion component in all five cases, together with evidence for a boundary effect on the low b-value portion of the signal intensity curve (rapidly attenuated fast diffusion component) in 4 of the 5 cases.\n\n*10−3 mm2 s−1 × (voxel size); **image intensity units × (voxel size). (Note the inverse distance dependencies in equation 6 and equation 8.)\n\nRegression coefficient posterior median estimates are listed together with the 0.025 and 0.975 posterior quantiles, which are given in brackets. β is the coefficient in the term that captures the boundary inverse-distance dependence of each of the 3 (phase, read and slice) voxel-specific diffusion coefficients, while α is the coefficient in a term that allows for boundary inverse-distance dependence in the b0 signal intensities. The regression model details are given in the Methods section.\n\nA visual representation of the magnitude of the boundary effect is given in Figure 3. The upper row shows the voxel-specific median ADCs estimates plotted against distance from the boundary. The corresponding ADC0.5 plots are given in the lower row. Superimposed on each graph is a median curve showing the boundary-distance dependence in ADCs, as given by the β/l term in equation 8. These curves have an arbitrary intercept (because the expression for diffusion includes additional random effect and intercept terms) and are shown with an intercept chosen to give a mid point equal to the overall median. The scatter in ADCs values at a given distance is attributable to the diffusion coefficient spatial random effect terms and provides a visual indication of the magnitude of the underlying 2D spatial heterogeneity in the diffusion coefficients (slow diffusion component). Consistent with the results given in the preceding section, and the statistics listed in Table 4, these plots show that the boundary distance effect makes a marked contribution to the spatial dependence in ADCs and that it is not completely dominated by the 2D spatial heterogeneity captured by the diffusion coefficient random effect terms. The relationship between ADC0.5 and β/l is less obvious indicating the importance of other sources of spatial variation. Scatter in the ADC0.5 values at a given distance from the boundary is accounted for by the entire set of random effect terms that are incorporated into the model, including those that capture the 2D spatial heterogeneity in the fast diffusion component.\n\nVoxel-specific ADCs estimates (top row) and ADC0.5 estimates (bottom row) in the boundary region plotted against distance from the boundary (distance given in units equal to the voxel size). ADCs is the slow ADC component, while ADC0.5 is derived from the half-maximum signal intensity estimate, and captures the signal attenuation that occurs at low b-value in addition to the slow component. The voxel-specific posterior medians are shown as dots while the posterior 0.025 and 0.975 quantiles are shown as bars. Although the figure does not show which of the medians belongs to each pair of quantiles, it does serve to provide an indication of the between-voxel differences in ADC together with the uncertainty in the estimates. Each plot includes a curve showing the dependence given by the inverse-distance diffusion term (β/l in equation 8), each of which is plotted with an intercept chosen to give a mid point equal to the overall median ADC value.\n\nA comparison of the data shown in Figure 3 with those given in Table 3 reveals small differences between the two, particularly with respect to the ADC0.5 maxima, some of those listed in the table being noticeably larger than shown in the figure. In fact, all of the minimum and maximum ADC0.5 values listed in the table are more extreme than shown in the figure, although in some cases the difference is negligible. Focusing on the estimated maxima, the differences arise because the figure shows the median voxel-specific ADC0.5 estimates (and quantiles) while the table lists the median (with 95% posterior interval) of the maximum ADC0.5 estimates. The latter are not voxel-specific, but take account of uncertainty in the voxel associated with the maximum values. The median of the maximum ADC0.5 estimates tend to be larger than the maximum of the median voxel-specific estimates, which is expected. Similar differences arise in the ADCs results, but are less obvious. The estimates given in the tables provide the required characterisation of tumour heterogeneity. On the other hand, Figure 3 provides a visual impression of the heterogeneity among the voxel-specific ADC estimates, and the dependence on boundary distance. Restated, the maximum of the voxel-specific median ADC values does not equal the median of the maximum ADC estimates, within the selected region. It is the latter that are used to derive the ADC range statistics, because these capture uncertainty in the identity of the voxels responsible for the extreme values. The capacity to account for all sources of variation/error is among the advantages of the MCMC modelling approach adopted in this study. Among the limitations of a standard independent-voxels analysis (i.e., one based on independent voxel-specific estimates) is the lack of a formal mechanism for achieving this.\n\nFinally we focus on the spatial-CAR precision parameters associated with the three diffusion coefficients (phase, read and slice-direction coefficients). They determine the magnitude of the dispersion in the diffusion coefficient spatial random effects, i.e., the underlying 2D spatial heterogeneity in ADCs. The boundary and core-region parameter estimates are listed in Table 5, after conversion to standard deviations in order to show these on the same scale as the ADC estimates. Inspection of these data suggests the presence of considerable anisotropy in the level of diffusion heterogeneity, together with substantial between-subject differences. The values obtained for the core region in Subject 2 are particularly large, raising questions regarding the robustness of these statistics. We return to this issue in the Discussion.\n\n*The posterior median estimate of each standard deviation is listed together with the 0.025 and 0.975 quantiles, which are given in brackets.\n\nCAR precision parameter values are listed after conversion to standard deviations. These provide a measure of local spatial dispersion and are shown on the same scale as the ADCs estimates. Specifically, the CAR precision (inverse variance) parameters determine the magnitude of local dispersion in the spatial random effects, i.e., the level of 2D spatial heterogeneity in the gradient-specific diffusion coefficients. It should be noted that the CAR precision parameters are conditional (by definition), and that a direct comparison with the other measures of dispersion given in this paper is invalid. Restated, measures of local dispersion are not, in general, directly related to global variability.\n\nIt should be noted that the standard deviations reported in Table 5 (square root of the inverse CAR precision parameter values) are not necessarily interpretable because they relate to local spatial structure as opposed to global variation over the entire ROI. Thus, in general, it is not meaningful to compare these conditional (local) dispersion parameters with standard measures of dispersion (see the Discussion for additional comments). As it happens, we do observe a relationship between each of the precision parameters and the corresponding region-specific diffusion-coefficient range of values. But this relationship is not guaranteed. We include Table 5 only because these data appear to suggest a considerable anisotropy in spatial structure.\n\nConvergence to a stationary distribution is a critical requirement in any statistical modelling analysis performed using MCMC. In the absence of convergence the resulting parameter estimates can be meaningless. In accordance with accepted procedure, an initial convergence assessment was performed using overlaid parallel-chain trace plots. Given a set of chains started at overdispersed positions in parameter space, a failure to achieve a good coverage of the region of parameter space supported by the posterior distribution is usually revealed by visual inspection. An additional semi-formal analysis was performed using the diagnostic tests listed in the Methods section. These tests were mainly restricted to the derived heterogeneity measures that are the focus of the study, including α, β, the boundary- and core-region ADCs range variables, the boundary- and core-region ADC0.5 range variables, a derived parameter equal to the boundary spatial range with the β/l contribution removed, and the CAR precision parameters. In addition a few voxel-specific ADCs parameters were also examined. The resulting convergence test and simulation accuracy results are given in Note S2. In summary, there were no instances of compromise due to convergence failure. Regarding simulation accuracy, Raftery-Lewis calculations indicated that 5000 samples (after thinning) were more than sufficient to provide the majority of nominal 95% credible intervals with a true coverage of between 94% and 96%, with probability 0.95. Where this was not achieved, the results indicate a true coverage of between 93% and 97%, with probability 0.95. We regard this level of accuracy to be satisfactory, noting that the heterogeneity statistics provided in this paper are generated after combining the three individual chains generated for each case, thus providing an accuracy greater than given here for the individual chains.\n\nAs stated previously, convergence and simulation accuracy are not the only considerations. Clearly, model adequacy is central to the present analysis because an incapacity to capture the true DWI signal intensities is expected to give rise to meaningless measures of ADC heterogeneity. Posterior predictive simulation was used to examine the MCMC output for signs of model inadequacy. It provides a mechanism for distinguishing between observations that are unexpected under the model (i.e., indicative of model failure), and observations that are compatible with the model despite appearing extreme. Details are given in Note S2. Although there are instances where a very low Bayesian p-value was obtained, this is not unexpected given the relatively large number of signal-intensity observations involved in each image dataset. Paying attention to the Bayesian p-values obtained for those voxels that give rise to the extreme parameter values that determine each of the reported heterogeneity estimates (i.e., the various range statistics), there are no instances where the p-value falls below 0.01. The conclusion is that there are no instances where an extreme ADC estimate gives rise to a spurious measure of heterogeneity, i.e., where an inflated heterogeneity estimate arises due to a spurious signal intensity observation or due to model failure. Re-stated, the model appears adequate in terms of its capacity to capture the true underlying tumour ADC heterogeneity.\n\n\nDiscussion\n\nThe focus of this study is a statistical modelling procedure for characterising intra-tumour heterogeneity. This was motivated by a well-established literature indicating that tumour heterogeneity has major implications for the development of improved treatment strategies and for the basic understanding of tumour development and pathophysiology. Among the important features of the approach that has been adopted is a single-stage analysis in which spatial heterogeneity is modelled simultaneously with signal intensity fitting. This is achieved by using a Bayesian spatial random effects model, implemented using MCMC. Some might question the need to adopt a formal spatial model, as opposed to a standard independent voxels (i.e., voxel-by-voxel) analysis. A brief statement of the general advantages of random effects modelling over an independent-units analysis is given in the Section on Bayesian random effect models that is included in the Introduction. In particular, we refer to the improvement in parameter precision that arises due to so-called information borrowing. Among the main ingredients of the random effects treatment adopted in this study are the spatial distributional constraints imposed by the CAR priors. These introduce a degree of spatial smoothing, referred to as shrinkage in the random effects context. The resulting voxel-specific ADC estimates and range statistics will, in general, be less extreme than those obtained from an independent-voxels analysis. In particular, the various range estimates are expected to be smaller than those obtained by subtracting the voxel-specific maximum and minimum ADC values obtained in an independent-voxels analysis. In that sense, the present random effects model analysis yields robust/conservative measures of heterogeneity.\n\nThe tumour heterogeneity analysis yields a variety of spatial statistics that are potentially useful from a clinical perspective. The main conclusions are that the previously reported relationship between diffusion and boundary proximity remains observable and achieves statistical significance (the 95% posterior intervals exclude zero) after adjusting for an underlying random 2D spatial heterogeneity in the diffusion model parameters. In addition, the results suggest that the tumour core and boundary regions are distinguishable in terms of the ADC spatial structure captured by the random effect terms, notwithstanding the additional deterministic boundary effect. A comparison of the magnitude of the deterministic boundary distance-effect with the underlying random 2D boundary heterogeneity suggests that both are important sources of variation in the vicinity of the boundary. No consistent pattern emerges from a comparison of the boundary and core spatial heterogeneity. In particular, the level of heterogeneity in the core is not consistently less than that observed in the boundary region, despite the additional boundary distance-effect contribution to the latter.\n\nA potentially interesting observation is that the diffusion coefficient CAR-prior precision parameters (these conditional parameters are measures of local heterogeneity) exhibit a significant degree of anisotropy, which suggests that combining the direction-specific diffusion coefficients to obtain ADC measures of heterogeneity is accompanied by a loss of information. But caution is required given the possibility of over-interpretation due to over-fitting the data which consists of only seven signal intensity observations per voxel. The present study suffers a deficiency that is not uncommon in the MRI field, namely existing imaging data acquired using a standard clinical imaging sequence were used in a retrospective modelling exercise. The DWI acquisition protocol was designed to produce clinical ADC maps sufficient for visual inspection. Data of this type are unlikely to be optimum from a modelling perspective. In the present context a lack of replication within each dataset, especially the lack of b0 replication, is particularly problematic. Obviously, the restriction to two non-zero b-values is an additional and severe limitation, noting the need to cater for departure from mono-exponential behaviour. Given signal intensity data consisting of 7 observations per voxel (b0, b500x, b500y, b500z, b1000x, b1000y, b1000z) the calculations are barely tractable. Clearly, b0 data replication would provide a more robust estimate of the true b0 signal intensity, thus facilitating estimation of the magnitude of the departure from mono-exponential behaviour and the noise contribution to the b0 signal intensity observation. Given the present data, this is achieved only through the combined constraints provided by the error distribution and the autoregressive spatial prior associated with the b0 signal-intensity observation. The residual error variance would be estimated with improved precision given a reasonable amount of replication.\n\nAn additional compromise arises in relation to the ADC0.5 statistics, because these are based on an estimate of the b-value at half-maximum signal intensity which, for some voxels, lies below 500 s mm–2. Clearly it would be preferable to acquire data using a protocol that gives a better coverage of the critical range of b-values, and with suitable replication. Scan time is, however, a limiting factor in achieving this ideal. Furthermore, replication of the entire DWI dataset and/or the entire imaging session would allow a formal assessment of the robustness of the various heterogeneity measures, and the suggestion of anisotropy in the CAR precision parameters. Unfortunately, given the need to work in a standard clinical setting, comprehensive within-scan replication coupled with replicate scanning is not a realistic option. In the absence of replicate data, a question arises regarding the plausibility of some of the dispersion statistics reported in this paper, in particular the larger of the standard deviations given in Table 5. Although the latter conditional dispersion statistics cannot be interpreted as measures of global heterogeneity, some might appear greater than expected. Nevertheless, these are not incompatible with the ADC data provided by Bull et al., (2012), as obtained by averaging over the entire tumour44. They report that subject-specific PNET average ADC values lie in the interval 0.67 × 10–3 to 1.23 × 10–3 mm2 s–1, based on an examination of 22 cases.\n\nAs originally conceived, the main purpose of a simultaneous spatial modelling analysis of core and boundary regions was improved parameter estimation. An initial working assumption was that the underlying random 2D spatial heterogeneity in the core and boundary regions would be similar, with the boundary-distance effect superimposed close to the boundary. Given a set of common spatial parameters, the information provided by the core would lead to improved precision in the boundary-specific parameter estimates. It became immediately obvious, however, that this preliminary assumption was wrong and that the level of 2D spatial correlation in the core and boundary regions is distinguishable, regardless of the additional boundary-distance terms that were included in the model. The model was modified accordingly. Although the more general model does not offer the advantage of improved boundary parameter estimation based on information borrowed from the core, the precision of the resulting parameter estimates is nevertheless sufficient. In particular, a characterisation of the difference between the core and boundary regions is provided by the modified model, in addition to an estimate of the magnitude of the boundary-distance effect. In summary, the core and boundary regions differ in their spatial correlation structure, requiring our initial model to be modified through the inclusion of region-specific spatial random effect terms. Although this complicates the comparison of the boundary-distance effect with the underlying random 2D spatial heterogeneity, it is possible to obtain sufficiently precise estimates of the magnitude of these two sources of spatial variation.\n\nA comment is required concerning the partial volume problem that arises in MRI due to finite resolution. Related issues include point spread function and zero filling effects. These must impact on the boundary distance coefficient estimates obtained in this study. For example, given the extreme case in which a step change occurs at the tumour boundary, partial volume/finite resolution effects, combined with image processing distortions, will cause a dispersion of the underlying step change in tissue characteristics. This interesting issue is related to the distinction between modelling the image-intensity data and modelling the underlying tissue structure. In keeping with standard practice among analysts engaged in MRI post-processing work, the results presented in this paper are based on modelling the image intensity data. An alternative approach might be sought in which a latent variables model is constructed, aimed at capturing the unobserved underlying tissue structure, combined with a model for the point spread function and zero-filling effects. A latent variables model must deal with all sources of image degradation. The resulting model will be complicated, however, and this will cause parameter estimation/precision problems, especially when working with sparse DWI data. A second point relates to the prognostic modelling literature that motivated this study. As stated in the Introduction, it has been suggested that PNET patient survival is a function of the tumour-boundary ADC gradient. Even in the extreme case in which a step change in tissue structure occurs at the tumour boundary, the magnitude of the gradient derived from the DW image will be related to the magnitude of the underlying step change. In particular, the magnitude of the ADC gradient will tend to zero as the step change tends to zero, and the relationship between them is expected to be monotonic. For this reason it is reasonable to assume that patient outcome will remain a function of any sensible regression coefficient derived from the tumour boundary DWI data, despite the degradation cause by imaging constraints and data processing, if a relationship genuinely exists between the underlying structure and survival. Existing literature suggests that the DWI data carry prognostic information, despite image degradation. We do acknowledge that a comparison of a given boundary-distance coefficient and the various measures of 2D heterogeneity is compromised, if the former is interpreted as a direct indicator of real underlying structure. For those readers preferring to dismiss the observed boundary decay in ADC as largely artefactual, caused by image degradation, we add the following rejoinder. The fact remains that the statistical model must include one or more terms to deal with the boundary effect, even if it is an imaging/data-processing artefact. The boundary-distance coefficient will be a function of the magnitude of the assumed step change that occurs at the boundary, coupled with imaging effects. From this perspective, the boundary distance-effect statistics listed in Table 2 and Table 4 might be regarded as measures of the magnitude of the imaging artefact, relative to the true 2D spatial heterogeneity in the region of the tumour boundary. We wish to stress that the notion of a tumour ADC dependence on boundary distance is not ours, and the main purpose of this study was to determine whether the boundary effect disappears after adjusting for an underlying 2D spatial heterogeneity in ADC. The present statistical analysis is not compromised by the possibility that the ADC boundary-distance effect is partly artefactual. The statistical model applies to the image data as opposed to the underlying structure. As it happens, the distance effect remains statistically significant, regardless of its origin. An additional note is warranted. This paper does not address any issues arising from the assertion that tumour ADC provides a reliable biomarker of cellularity, apart from the statements made in the Introduction regarding the limitations of simple correlation analyses and the criteria that should be met before claiming to have a reliable surrogate marker.\n\nPosterior predictive simulation was performed as a mechanism for assessing model adequacy. In essence this is a simulation approach to comparing an observed statistic, designed to capture some key feature of the data, with that given by the model (chapter 6 in 23,41–43). Bayesian p-values are commonly adopted as a measure of discrepancy between an observed statistic and that obtained under the model. We note that the approach has received some criticism. For example, the predictive probabilities are not calibrated (in general, the posterior predictive p-values do not have a uniform distribution under the null hypothesis45). Some data analysts have suggested that the very notion of Bayesian p-values is a contradiction. (An indication of the nature of the paradox is given on page 87 of 28.) A number of eminent statisticians remain enthusiastic about this approach to model evaluation, however. The early BUGS documentation46 included a section on goodness-of-fit tests based on Bayesian p-values. Gelman, who is a notable advocate, argues that a statistical model can seldom be perfectly true (see page 158 in 23, or page 776 in 42), but that it is important to demonstrate that it is adequate for the intended purpose, even if it is deficient in some other aspect. Posterior predictive simulation provides a useful tool for performing this kind of model assessment. Nevertheless, we acknowledge that controversy remains regarding some aspects of this approach, including the calibration issue referred to above and an uncertainty regarding the p-value threshold that is used to indicate a problem. Thus the simple analysis adopted in the present study is undertaken without reference to the expected distribution under a satisfactory model. The purpose is to verify that there are not too many instances where an observation appears unexpected under the model. In the present study it was used as an approach for detecting potentially spurious signal intensities and/or instances of model failure, and to determine whether these are associated with extreme ADC estimates. We found two individual signal intensity observations with extremely low p-values. Given the nature of DWI and its sensitivity to movement, spurious observations might be expected. Thus, extremely low Bayesian p-values might be attributed to the simplicity of the error term, which ignores the possibility of spurious observations caused by motion and other imaging problems, as opposed to an inadequacy in the deterministic and/or spatial components of the model. That said, these two DWI observations did not give rise to extreme ADC estimates. Thus, exaggerated spatial heterogeneity measures arising from spurious DWI observations do not appear to be a problem. Accordingly, we conclude that the heterogeneity statistics given in this paper are robust to the presence of outlier signal intensities. Nevertheless, some might take the view that some form of model refinement should have been undertaken in an attempt to deal with the occurrence of a number of small Bayesian p-values. As a rejoinder we would argue that model assessment requires more comprehensive data than that provided by a standard clinical DWI acquisition. As stated previously, replication is desirable at several levels, including repeated acquisition within individual DWI datasets (as opposed to signal accumulation) and within-session DWI dataset replication. Given a reasonable level of replication, model refinement based on residuals analysis and other criteria becomes realistic. Furthermore, as stated above, replication at the DWI-dataset and/or imaging-session level would also facilitate an assessment of the robustness of the results.\n\nDespite the limited number of observations per voxel and the resulting compromises, the signal intensity residuals were examined for signs of model inadequacy, as a complement to the assessment that was performed using posterior predictive simulation. In addition to a small number of relatively large residuals which are attributable to spurious signal intensity observations, the residual plots do display a degree of remaining spatial dependence. This does indicate a degree of model inadequacy, including a potential deficiency in the form of distance dependence that was adopted and/or the assumptions underlying the random effects. For example, the present random effects treatment is based on spatially invariant CAR precision parameters (these determine the level of local smoothing), and this might be an oversimplification. A spatially adaptive model might be investigated, although this is not a trivial undertaking. Alternatively, the error term might be modified to capture the remaining spatial structure, thus dropping the independent residuals assumption and substituting some form of autoregressive error behaviour. In order to adopt the latter approach as a sensible solution, the magnitude of the residuals must remain small relative to the total spatial variation in signal amplitude, thus ensuring that the majority of the spatial variation is captured by the random effect and boundary effect terms. Restated, the spatial heterogeneity estimates derived from the analyses will be compromised if a substantial proportion of the intra-tumour variation is captured by the error term. A heavy-tailed error distribution might provide a mechanism for dealing with spurious observations arising from a sensitivity to motion. Clearly, the model that was adopted in this study cannot be regarded as definitive. As stated above, model assessment and refinement, including an examination of the assumptions underlying the error term, would be facilitated by data replication. In particular, a comprehensive dataset with replication would permit a meaningful examination of alternative boundary decay models.\n\nA final comment on the form of the model used in this study relates to the decision to adopt CAR priors for the spatial random effect terms. As stated in the Results section, the CAR prior precision parameters cannot be used as direct measures of global tumour heterogeneity because these relate to local spatial correlation structure as opposed to global structure. Trial analyses were performed using a global model of spatial heterogeneity based on so-called exchangeable priors. This would have offered the advantage of providing more direct measures of heterogeneity. Unfortunately, this model tended to produce poorly distributed residuals due to over-fitting, a problem that might be expected given the lack of replicate signal intensity observations and poor coverage of the b-value range. For this reason, given the present data, models based on exchangeable priors were abandoned as a potential alternative to the present CAR-prior models.\n\nThe preceding qualifications regarding the validity of the spatial model and/or error term prompt us to make a final comment regarding the value of this study. The question is whether it was sensible to embark on a study using clinical data that are sub-optimum from a modelling perspective. In our view, the paediatric data available to us represent a valuable and rare resource, despite the limitations arising from the acquisition constraints of a standard clinical imaging environment. We suggest that using these data in an exploratory study is justified, and provides a valid mechanism for gaining insight into the utility of the information that might be derived from these data. There is no possibility of re-scanning these children using an experimental imaging protocol for the sole purpose of undertaking an exploratory study of the potential benefits of a given kind of analysis. The results obtained from this preliminary study using existing data gives an indication of the possibilities, enabling a decision to be made regarding the development of this approach. A clear indication of the manner in which the imaging protocol might be improved also emerges, although any proposed changes are subject to the constraints that inevitably arise in a clinical setting. We have shown that despite the limitations of an analysis based on standard clinical DWI data (mainly a lack of replication), the heterogeneity summary measures have sufficient precision to be useful. The spatial model does achieve a separation of the noise contribution from the effects of departure from mono-exponential dependence, despite the absence of replication in the b0 signal intensity observation. The constraints imposed by the CAR spatial prior and the spatially invariant error distribution render the problem tractable. The key is a simultaneous modelling of a collection of voxels, since the separate estimation of the true signal intensity and noise contributions to a single b0 observation is impossible in an independent-voxels analysis. A comparison of the ADC estimates (both ADCs and ADC0.5) obtained with the current spatial model and a simple voxel-by-voxel fitting of the DWI signal intensity data (treating the b0 observations as error free) indicates marked differences in some voxels, which is expected and attributable to the shrinkage/smoothing properties of the random effects model.\n\nAnother matter that requires attention is ROI selection. In this study boundary ROIs were positioned where the tumour border is very well defined, working under the assumption that boundary effects might be more pronounced in that region. More sophisticated approaches should be investigated if the models outlined in this paper are to be adopted for prognostic modelling. Apart from the practical issues of tumour segmentation and automation, a modified approach will be required to cater for the expectation that tumour boundary heterogeneity is itself position dependent. Averaging heterogeneity over the entire boundary might obscure important prognostic information if, for example, tumour evolution is not uniform over the boundary. It would not be surprising to find improved performance among survival/prognostic models that retain position dependent heterogeneity information, compared with those based on averaged data or data taken from arbitrary regions.\n\nIn summary, the present study suggests that heterogeneity measures can be derived from standard clinical DWI datasets, despite their limited information content. Particular attention is paid to heterogeneity close to the tumour boundary, in order to determine whether water diffusivity in the boundary region of some tumours exhibits a deterministic dependence on distance from the boundary. The results indicate that the boundary-distance effect retains statistical significance after adjusting for an underlying and general 2D spatial heterogeneity. The level of spatial heterogeneity in the region of the boundary is not consistently greater than that observed in the core. The analysis could be extended to determine whether the heterogeneity parameters provide useful prognostic indicators in a survival analysis. Obviously, the same approach could be adopted using data acquired with a purpose-designed sequence, the advantage being that an increase in the accuracy and precision of the heterogeneity measures will be an advantage if these do carry useful prognostic information. In addition to assessing these heterogeneity measures as useful predictors in a survival analysis, the question arises regarding the relationship between these measures and intra-tumour genetic spatial heterogeneity. A biomarker of genetic heterogeneity would provide a powerful tool with applications in both patient management and in cancer research. Clearly, any imaging method that fulfils this role has the potential to provide clinical insights relevant to individual treatment and the pursuit of a better understanding of cancer biology.\n\n\nData availability\n\nF1000Research: Dataset 1. Diffusion-weighted MR signal intensity observations and voxel adjacency data. 10.5256/f1000research.9355.d13281648",
"appendix": "Author contributions\n\n\n\nMDK conceived, designed and performed the analyses and wrote the manuscript. MG-S collated and selected the data, read the manuscript and suggested improvements.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors thank Prof D. Gadian for reading early drafts of this paper and making substantial suggestions, each leading to marked improvements. Matthew Grech-Sollars thanks Prof Chris A. Clark and Prof Andrew C. Peet for support received as a participant in the research undertaken by the Children’s Cancer and Leukaemia Groups (CCLG) Functional Imaging Group. Martin King thanks Prof Chris A. Clark for covering the costs of attending a number of workshops held by the CCLG Functional Imaging Group. Various presentations given at these workshops provided the initial motivation for the statistical modelling work outlined in this paper.\n\n\nSupplementary Information\n\nMarkov random field models and the conditional autoregressive prior Markov random field models are well known among fMRI researchers, but a brief overview follows, as it applies to the present study. Markov processes are particularly prominent in time-series data analysis. In that setting a first order Markov process is defined as a stochastic process in which the conditional probability of some future state is determined by the present state, unaffected by past history. Thus the process is first order autoregressive (AR(1)) given by Xt = αXt–1 + Zt (see, for example, page 35 in 47) where Zt is some random process with zero mean (Xt is the autoregressive process). An extension of this kind of conditional independence to the spatial case leads to Markov random field models in which the probability distribution of a random variable at some position in space is completely determined by conditioning on a set of neighbouring values. Thus a Markov random field characterisation of a spatial process is based on the assumption that the conditional distribution of some variable at a given location depends only on the value of this variable at a subset of immediately neighbouring locations. In the present context, this is the prior assumption, and it is modified by the information provided by the data, as expressed in the likelihood. The magnitude of the estimated differences between neighbouring voxels, ie., the extent to which the voxel-specific estimates are shrunk towards less extreme values, depends on the information content of the data, relative to the prior.\n\nAn intuitive Markov random field model, widely used as a prior distribution, is the intrinsic Gaussian conditional autoregressive (CAR) model which, for some parameter U, takes the form (see, for example, 24–26,31)\n\nConvergence tests, simulation accuracy, posterior predictive simulation and Bayesian p-values\n\nConvergence tests were performed as outlined in the Methods section. There were no instances in which the Gelman-Rubin diagnostic suggested convergence failure. The majority of the Geweke Z-scores were also satisfactory, although several chains yielded a Z-score above 2. In only one case was the Z-score extreme (a value of 4.4 was obtained for one of the 3 chains of the boundary-region ADCs range parameter in Subject 1). Despite these indications of some non-ideal Z-scores, a comparison of the parameter median estimates and 95% posterior intervals given by the individual chains indicated good agreement. Thus, we conclude that there is no instance of unreliable measures of heterogeneity caused by compromised convergence.\n\nThe Raftery-Lewis results indicated that in many cases 5000 samples (after thinning) were more than sufficient to provide the 0.025 quantile estimates with an accuracy of ±0.005 with probability 0.95. Thus the resulting nominal 95% credible intervals have a true coverage of between 94% and 96%, with probability 0.95. There were, however, many other instances where this level of simulation accuracy was not achieved. Nevertheless, in these cases, the accuracy was at least ±0.01 with probability 0.95, which provides nominal 95% credible intervals with a true coverage of between 93% and 97%, with probability 0.95. It should be noted that the heterogeneity statistics provided in this paper are generated after combining the three individual chains, and that the resulting accuracy will be greater than given here for the individual chains.\n\nModel adequacy is of central importance to the present analysis because an incapacity to capture the true DWI signal intensities is expected to give rise to meaningless measures of ADC heterogeneity. There were several instances where an observed DWI signal intensity appears extreme and the corresponding signal intensity estimate exhibits shrinkage towards a more typical value. This behaviour is expected under a random effects model, and is not necessarily an indication of model failure. Posterior predictive simulation provides a mechanism for distinguishing between observations that are unexpected under the model (i.e., indicative of model failure), and observations that are compatible with the model despite appearing extreme. To this end, Bayesian p-values were calculated for every observation. The focus of this paper is tumour heterogeneity, as characterised by the extremes in ADC0.5 and ADCs. Thus, particular attention is paid to the Bayesian p-values obtained for those voxels that give rise to the extreme parameter values that determine each of the reported heterogeneity estimates (i.e., the various range statistics), noting that extreme signal-intensity observations do not necessarily give extreme ADC values, and vice versa. In three of the five cases, no observation yielded a Bayesian p-value less than 0.001. The exceptions were Subject 2 and 4, which, taken together, yielded a total of three p-values of approximately 6 × 10–4. These 2 cases yielded a number of additional observations with a p-value < 0.05, as did the other 3 cases. Given the relatively large number of signal-intensity observations in each of the image datasets this is expected regardless of model adequacy. The largest number of observations with low p-values was obtained for Subject 1, with 20 p-values in the range 0.001 < p-value < 0.01. Inspection of the raw signal intensity data and diffusion weighted images indicates that, in this particular case, a small image artefact contributes to the apparent discrepancy between the model and observed data. Focusing on those voxels with extreme ADC0.5 and ADCs values, as expected some voxels yield Bayesian p-values in the tails of the distribution (p-value < 0.05), but there are no instances where the p-value falls below 0.01. This is consistent with the conclusion that the estimated ADC extremes are not invalid, i.e., attributable to the excessive influence of spurious signal intensity observations or model failure. In contrast, the model appears adequate in terms of its capacity to capture the true underlying tumour ADC heterogeneity. The low Bayesian p-values obtained for some observations do suggest, however, that potential improvements to the model might be sought. An inspection of various plots (not shown) indicates the presence of some spatial structure in the residuals. Thus, despite the apparent complexity of the random effects model adopted in this study and the limited number of observations, just seven per voxel, the low Bayesian p-values obtained for some observations might indicate scope for model refinement. The Discussion takes up the model refinement issue.\n\n\nReferences\n\nMarusyk A, Almendro V, Polyak K: Intra-tumour heterogeneity: a looking glass for cancer? Nat Rev Cancer. 2012; 12(5): 323–334. PubMed Abstract | Publisher Full Text\n\nYap TA, Gerlinger M, Futreal PA, et al.: Intratumor heterogeneity: seeing the wood for the trees. Sci Transl Med. 2012; 4(127): 127ps10. PubMed Abstract | Publisher Full Text\n\nGerlinger M, Swanton C: How Darwinian models inform therapeutic failure initiated by clonal heterogeneity in cancer medicine. Br J Cancer. 2010; 103(8): 1139–1143. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSilva AS, Kam Y, Khin ZP, et al.: Evolutionary approaches to prolong progression-free survival in breast cancer. Cancer Res. 2012; 72(24): 6362–6370. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Sullivan F, Roy S, O’Sullivan J, et al.: Incorporation of tumor shape into an assessment of spatial heterogeneity for human sarcomas imaged with FDG-PET. Biostatistics. 2005; 6(2): 293–301. PubMed Abstract | Publisher Full Text\n\nO’Connor E, Fieller N, Holmes A, et al.: Functional principal component analyses of biomedical images as outcome measures. J R Stat Soc Ser C Appl Stat. 2010; 59(1): 57–76. Publisher Full Text\n\nChenevert TL, Stegman LD, Taylor JMG, et al.: Diffusion magnetic resonance imaging: an early surrogate marker of therapeutic efficacy in brain tumors. J Natl Cancer Inst. 2000; 92(24): 2029–2036. PubMed Abstract | Publisher Full Text\n\nGauvain KM, McKinstry RC, Mukherjee P, et al.: Evaluating pediatric brain tumor cellularity with diffusion-tensor imaging. AJR Am J Roentgenol. 2001; 177(2): 449–454. PubMed Abstract | Publisher Full Text\n\nHall DE, Moffat BA, Stojanovska J, et al.: Therapeutic efficacy of DTI-015 using diffusion magnetic resonance imaging as an early surrogate marker. Clin Cancer Res. 2004; 10(23): 7852–7859. PubMed Abstract | Publisher Full Text\n\nJenkinson MD, du Plessis DG, Smith TS, et al.: Histological growth patterns and genotype in oligodendroglial tumours: correlation with MRI features. Brain. 2006; 129(Pt 7): 1884–1891. PubMed Abstract | Publisher Full Text\n\nMatsumoto Y, Kuroda M, Matsuya R, et al.: In vitro experimental study of the relationship between the apparent diffusion coefficient and changes in cellularity and cell morphology. Oncol Rep. 2009; 22(3): 641–648. PubMed Abstract | Publisher Full Text\n\nYamashita Y, Kumabe T, Higano S, et al.: Minimum apparent diffusion coefficient is significantly correlated with cellularity in medulloblastomas. Neurol Res. 2009; 31(9): 940–946. PubMed Abstract | Publisher Full Text\n\nSugahara T, Korogi Y, Kochi M, et al.: Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging. 1999; 9(1): 53–60. PubMed Abstract | Publisher Full Text\n\nThompson G, Cain JR, Mills SJ, et al.: Apparent diffusion coefficient measures on MR correlate with survival in glioblastoma multiforme. Proc Intl Soc Reson Med. 2009; 17: 280. Reference Source\n\nPrentice RL: Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med. 1989; 8(4): 431–440. PubMed Abstract | Publisher Full Text\n\nSedgwick P: Pearson’s correlation coefficient. BMJ. 2012; 345: e4483. Publisher Full Text\n\nJenkinson MD, Smith TS, Brodbelt AR, et al.: Apparent diffusion coefficients in oligodendroglial tumors characterized by genotype. J Magn Reson Imaging. 2007; 26(6): 1405–1412. PubMed Abstract | Publisher Full Text\n\nKotsenas AL, Roth TC, Manness WK, et al.: Abnormal diffusion-weighted MRI in medulloblastoma: does it reflect small cell histology? Pediatr Radiol. 1999; 29(7): 524–526. PubMed Abstract | Publisher Full Text\n\nHamstra DA, Chenevert TL, Moffat BA, et al.: Evaluation of the functional diffusion map as an early biomarker of time-to-progression and overall survival in high-grade glioma. Proc Natl Acad Sci U S A. 2005; 102(46): 16759–16764. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoffat BA, Chenevert TL, Meyer CR, et al.: The functional diffusion map: an imaging biomarker for the early prediction of cancer treatment outcome. Neoplasia. 2006; 8(4): 259–267. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEllingson BM, Malkin MG, Rand SD, et al.: Volumetric analysis of functional diffusion maps is a predictive imaging biomarker for cytotoxic and anti-angiogenic treatments in malignant gliomas. J Neurooncol. 2011; 102(1): 95–103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrech-Sollars M, Saunders DE, Phipps KP, et al.: Survival analysis for apparent diffusion coefficient measures in children with embryonal brain tumours. Neuro Oncol. 2012; 14(10): 1285–1293. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGelman A, Carlin JB, Stern HS, et al.: Bayesian data analysis. 2nd ed. Boca Raton, FL: Chapman & Hall/CRC; 2004. Reference Source\n\nWakefield JC, Best NG, Waller L: Bayesian approaches to disease mapping. In: Elliott P, Wakefield JC, Best NG, Briggs DJ editors. Spatial epidemiology: methods and applications. Oxford: Oxford University Press; 2000; 104–127. Publisher Full Text\n\nRichardson S, Thomson A, Best N, et al.: Interpreting posterior relative risk estimates in disease-mapping studies. Environ Health Perspect. 2004; 112(9): 1016–1025. PubMed Abstract | Free Full Text\n\nBest N, Richardson S, Thomson A: A comparison of Bayesian spatial models for disease mapping. Stat Methods Med Res. 2005; 14(1): 35–59. PubMed Abstract | Publisher Full Text\n\nGilks WR, Richardson S, Spiegelhalter DJ, editors: Markov chain Monte Carlo in practice. London: Chapman & Hall; 1996. Reference Source\n\nCarlin BP, Louis TA: Bayesian methods for data analysis. 3rd ed. Boca Raton, FL: Chapman & Hall/CRC; 2009. Reference Source\n\nLunn DJ, Thomas A, Best N, et al.: WinBUGS - A Bayesian modelling framework: concepts, structure, and extensibility. Stat Comput. 2000; 10(4): 325–337. Publisher Full Text\n\nLunn D, Spiegelhalter D, Thomas A, et al.: The BUGS project: Evolution, critique and future directions. Stat Med. 2009; 28(25): 3049–3067. PubMed Abstract | Publisher Full Text\n\nBanerjee S, Carlin BP, Gelfand AE: Hierarchical modeling and analysis for spatial data. 2nd ed. Boca Raton, FL: CRC Press; 2015. Publisher Full Text\n\nKing MD, Gadian DG, Clark CA: A random effects modelling approach to the crossing-fibre problem in tractography. Neuroimage. 2009; 44(3): 753–768. PubMed Abstract | Publisher Full Text\n\nKing MD, Crowder MJ, Hand DJ, et al.: Temporal relation between the ADC and DC potential responses to transient focal ischemia in the rat: a Markov chain Monte Carlo simulation analysis. J Cereb Blood Flow Metab. 2003; 23(6): 677–688. PubMed Abstract | Publisher Full Text\n\nKing MD, Calamante F, Clark CA, et al.: Markov chain Monte Carlo random effects modeling in magnetic resonance image processing using the BRugs interface to WinBUGS. J Stat Softw. 2011; 44(2): 1–23. Publisher Full Text\n\nSpiegelhalter D, Thomas A, Best N, et al.: WinBUGS user manual. MRC Biostatistics Unit, Institute of Public Health, Cambridge, and Department of Epidemiology and Public Health, Imperial College School of Medicine, London; 2003. Reference Source\n\nThomas A, Best N, Lunn D, et al.: GeoBUGS User Manual. 2004. Reference Source\n\nGelman A, Rubin DB: Inference from iterative simulation using multiple sequences. Stat Sci. 1992; 7(4): 457–472. Publisher Full Text\n\nKass RE, Carlin BP, Gelman A, et al.: Markov chain Monte Carlo in practice: a roundtable discussion. Am Stat. 1998; 52(2): 93–100. Publisher Full Text\n\nCowles MK, Carlin BP: Markov chain Monte Carlo convergence diagnostics: a comparative review. J Am Stat Assoc. 1996; 91(434): 883–904. Publisher Full Text\n\nPlummer M, Best N, Cowles K, et al.: CODA: Convergence diagnosis and output analysis for MCMC. R News. 2006; 6(1): 7–11. Reference Source\n\nGelman A, Meng X-L: Model checking and model improvement. In: Gilks WR, Richardson S, Spiegelhalter DJ, editors. Markov chain Monte Carlo in practice. London: Chapman & Hall; 1996; 189–201. Reference Source\n\nGelman A, Meng X-L, Stern H: Posterior predictive assessment of model fitness via realized discrepancies. Stat Sin. 1996; 6: 733–807. Reference Source\n\nBayarri MJ, Castellanos ME: Bayesian checking of the second levels of hierarchical models. Statist Sci. 2007; 22(3): 322–343. Publisher Full Text\n\nBull JG, Saunders DE, Clark CA: Discrimination of paediatric brain tumours using apparent diffusion coefficient histograms. Eur Radiol. 2012; 22(2): 447–457. PubMed Abstract | Publisher Full Text\n\nGelman A: Two simple examples for understanding posterior p-values whose distributions are far from unform. Electron J Stat. 2013; 7: 2595–2602. Publisher Full Text\n\nSpiegelhalter D, Thomas A, Best N, et al.: BUGS 0.5. Bayesian inference using Gibbs sampling manual. MRC Biostatistics Unit, Institute of Public Health, Cambridge; 1996. Reference Source\n\nChatfield C: The analysis of time series. An introduction. 4th ed. London: Chapman & Hall; 1989. Reference Source\n\nKing MD, Grech-Sollars M: Dataset 1 in: A Bayesian spatial random effects model characterisation of tumour heterogeneity implemented using Markov chain Monte Carlo (MCMC) simulation. F1000Research. 2016. Data Source"
}
|
[
{
"id": "16652",
"date": "21 Nov 2016",
"name": "David A. Porter",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an extremely well-written manuscript that makes a significant contribution to the methodology used to analyse quantitative diffusion imaging of tumours with MRI. There is a rapid growth in clinical interest in this application of diffusion imaging. However, many studies to date have performed a rudimentary analysis of the data without the rigour of an informed statistical analysis. The current study addresses this issue by introducing a sound methodology for performing these analyses. The authors draw upon an extensive knowledge of statistical methodology and provide clear explanations of the techniques used, together with an extensive list of background references, which are clearly linked to individual aspects of the proposed methodology. I have provided a list of detailed comments below, but I would only expect corresponding changes to the manuscript where this can be done with very minor changes to the text.\n\nAs addressed in the discussion, the analysis of the local spatial variation using the CAR model characterizes the behaviour of the image pixel signal intensities, rather than the underlying biological tissue properties directly. The authors acknowledge that the imaging point spread function (PSF) therefore plays a role in this part of the analysis. Indeed, for the single-shot EPI sequence used in this study there is typically a significant image smoothing effect due to T2* decay during the long readout time. In addition to the T2* dependence, the extent of this smoothing depends on the additional acquisition parameters, so it would be helpful if the following additional information were added to the methods: EPI echo spacing; parallel imaging acceleration factor; partial Fourier factor. In addition, it would also be relevant to specify whether any image-based interpolation was applied to the images, as this would also affect the local correlation between pixels. Such interpolation is sometimes applied to correct distortions in EPI images due to concomitant gradient terms or eddy currents. As the spatial extent of these effects can be estimated for the specific experimental case, it would also be helpful for a brief statement regarding the impact that these issue are likely to have in the current study.\n\nPage 4, Image Data Analysis: The images used were presumably modulus images. Although, this is standard practice for DWI, I suggest that this is made explicit because there are of course implications for the noise distribution. In relation to this point, it is not clear to me from the manuscript whether the Rician noise distribution from such images was considered during the analysis. In standard analyses, the resulting systematic bias at low SNR invalidates the standard exponential decay model used in equation (2). Please provide a statement regarding the effect of modulus image reconstruction on the analysis performed in this work.\n\nPage 5, Image Data Analysis: The rigorous statistical modelling in the study is somewhat undermined by the empirical process of defining the tumour extent and intra-tumour core and boundary regions. This is already acknowledged in the discussion as a limitation, requiring future work for clinical application. However, in the context of the current study, the statistical characterisation of the boundary effects relies on prospectively identifying a boundary region using arbitrary ADC thresholds. It is possible that the regression coefficients in Table 4 are heavily influenced by this manual procedure. This could be explored experimentally, or addressed by extending the existing discussion. Would it be an option in the future to define the boundary zone automatically within the analysis by identifying the region with the spatial dependence of ADC value assumed in the model?\n\nI am not convinced by the use of the ADC0.5 parameters, whose definition seems quite contrived and not very intuitive. Comparing the data in Tables 1 and 3, the differences between ADCs and ADC0.5 do however make sense: the ADC0.5 values tend to be higher due to the influence of the low b-value, fast-diffusion component, particularly for the Maximum values, which presumably correspond to voxels with relatively high fractions of the fast diffusion behaviour. However, I can imagine that this information would have been more accessible if it had been provided via a normalized version of the delta parameter in Fig. 2, which is a more direct measure of the deviation from monoexponential behaviour at low b-value.\n\nAs a point of reference, it would be interesting to see the corresponding results for spatial heterogeneity in regions of brain tissue with ADC values in the normal range or for brains of healthy volunteers. This might help to establish to what extent the large subject-dependent variation in spatial heterogeneity in the clinical data is meaningful.\n\nPage 13, Discussion: The authors state that the proposed MCMC method is expected to produce ADC range data that is less extreme than values provided by an independent voxels analysis. To substantiate this claim, it would have been helpful to provide a direct comparison of results generated by applying both methods to the current data set.\n\nThe field strength of scanner was omitted from the Methods (1.5T).\n\nPlease state in the Methods what type of diffusion preparation was used (Stejskal-Tanner or twice-refocused). I assume that the twice-refocused method was used to minimize diffusion-gradient-dependent image distortion (naturally, an important consideration in quantitative studies in general and specifically relevant to the analysis of local spatial variation in the current work).\n\nThe symbols used for signal offset at b=0 (equations 3,5,6) and the set of nearest-neighbour voxel indices (equations 12,13) are superficially similar and this might cause confusion.",
"responses": [
{
"c_id": "2918",
"date": "28 Jul 2017",
"name": "Martin King",
"role": "Author Response",
"response": "We are extremely grateful to Dr. Porter for the care and time he has taken to produce his report and for his helpful comments. Each point is well-founded and we hope to provide satisfactory changes and responses.Comment 1 ... it would be helpful if the following additional information were added to the methods: EPI echo spacing; parallel imaging acceleration factor; partial Fourier factor. In addition, it would also be relevant to specify whether any image-based interpolation was applied to the images. ... it would also be helpful for a brief statement regarding the impact that these issue are likely to have in the current study.Response to Comment 1. EPI echo spacing, 0.75ms; parallel imaging acceleration factor, 2; partial Fourier factor, 6/8; image-based interpolation, the original resolution was 128x128, increased to 256x256.Comment 1 cont. '...a brief statement regarding the impact that these issues are likely to have ...'Given the option, we would avoid interpolation and other preprocessing manipulations of the data, because it makes little sense to modify the correlation structure at one stage of an analysis and then, in a subsequent processing step, adopt a model designed to capture the correlation structure. It would be preferable to work with the raw data using a single-step processing method and an appropriate model, designed to capture the important features of the raw data. Restated, multistage analysis involving preprocessing steps is undesirable, as opposed to single stage analyses using a model designed to capture the true data structure, thus dealing with all sources of variation and other effects. It might be noted that we have (Discussion page 14-15), commented on the distinction between modelling the dMRI data (effected by point-spread, partial volume, preprocessing effects, etc.) versus modelling the underlying tissue structure.Focussing on interpolation effects, a common disadvantage of retrospective MRI studies involving the statistical analysis of routine clinical imaging data, is the lack of control over data collection and preprocessing. The fact is, however, that the preprocessed image data used in this study are all that was available. The manner in which interpolation impacts on the results presumably depends on the true underlying spatial structure and signal to noise ratio. The effects are expected to be case specific. We do not have the data required to explore the effects of interpolation.Bearing in mind that the primary objective was to determine whether the previously reported boundary distance effect remains significant, after adjusting for the underlying 2D spatial variation, we believe that interpolation effects should not be a serious issue. The demonstration that the boundary effect retains its significance in all 5 cases is unlikely to be an artefact of interpolation.Relevant parts of the Discussion could be extended, given Dr. Porter's comment. We already mention (page 14) various other disadvantages arising from the use of routine clinical data. We believe that a lack of replication is the major limitation. We point out that sub-optimum data is a common problem encountered by analysts engaged in MRI statistical modelling calculations. Often, the analyst has no option but to use existing image data acquired and processed using a standard clinical imaging method, designed to produce clinical maps/images deemed sufficient for visual inspection. These data are not expected to be optimum from a modelling perspective. Usually there is no possibility of collecting new data, optimised with statistical modelling in mind. Dr. Porter's comment on interpolation is clearly related to the distinction between modelling the MRI data versus modelling the underlying tissue structure. On page 15 we discuss the prognostic modelling implications of this distinction.Comment 2 (page 4), Image Data Analysis: The images used were presumably modulus images. Although, this is standard practice for DWI, I suggest that this is made explicit because there are of course implications for the noise distribution.Response to Comment 2. Modulus image data were used in the analyses. A revised Methods section should make this clear.Comment 2 continued; Rician noise. It is not clear to me from the manuscript whether the Rician noise distribution from such images was considered during the analysis. In standard analyses, the resulting systematic bias at low SNR invalidates the standard exponential decay model used in equation (2). Please provide a statement regarding the effect of modulus image reconstruction on the analysis performed in this work.Response to Comment 2 on Rician noise: We have not included a Rician error term but treated the residual error term as having a normal distribution (equation 1). In the Discussion, we acknowledge that our treatment of the residual error is rudimentary, and list several issues that should be addressed in order to improve the statistical model (page 16). In particular we discuss the fact that a residuals analysis suggests some deficiency in the model (a low level of residual spatial structure is observed, as opposed to independent residuals, randomly distributed about zero). Given the numerous sources of variation that impact on the spatial dMRI data (in addition to the NMR noise arising from the sample and coil), we do agree that improved modelling of the residual error distribution is required to take this forward, and suggested the need for a more comprehensive dataset with adequate replication to achieve this. Presumably one needs to distinguish between NMR noise, and its distributional effect on the magnitude data, versus the effect of the other sources of variation on the distribution of the fitted residuals. (For example, spurious signal intensity observations arising from imperfect magnetic field gradient sensitisation is a common feature of DWI data.) As stated above, we made the simplifying assumption that the residuals have a distribution that is reasonably approximated by the normal distribution. The usual argument for adopting a normal distribution for dealing with multiple additive random processes is to invoke the Central Limit Theorem. Furthermore, the b-values used to generate the clinical DWI data are relatively modest and the level of signal attenuation is not, therefore, excessive. In the Discussion we refer to the expected sensitivity to motion, and suggest the need to deal with spurious observations by adopting some form of heavy-tailed distribution, believing that this is a particularly pressing issue. In summary, given the many sources of variation/noise, the manner in which these accumulate is uncertain. Taking this forward we would need to determine the dominant form of the residual error distribution. The feasibility of adopting more complicated error distributions was not examined in the present study because the current data are lacking in both replication and design points (b-values). Given a more comprehensive dataset with adequate replication, an investigation of alternative residual error models, including heavy-tailed models (implemented using mixture error models, for example) would be desirable. Clearly Dr. Porter's comment on the Rician distribution should be considered in this context.Comment 3 (page 5). The study is undermined by the empirical process of defining the tumour extent and intra-tumour core and boundary regions. This is already acknowledged in the discussion as a limitation, requiring future work for clinical application. However, in the context of the current study, the statistical characterisation of the boundary effects relies on prospectively identifying a boundary region using arbitrary ADC thresholds. It is possible that the regression coefficients in Table 4 are heavily influenced by this manual procedure. This could be explored experimentally, or addressed by extending the existing discussion. Would it be an option in the future to define the boundary zone automatically within the analysis by identifying the region with the spatial dependence of ADC value assumed in the model?Response to Comment 3. Dr. Porter highlights the centrally important boundary identification issue. An earlier draft of this manuscript included additional text devoted to this topic, but this was removed to achieve a reduction in word count. We take this opportunity to clarify our position.We start by emphasising that the primary objective underlying this study was to examine the boundary distance effect, and to determine whether it is robust and remains significant after adjusting for the underlying 2D spatial heterogeneity in diffusivity. The tumour DWI boundary-distance effect has received much attention in the MRI literature. So the primary research question is whether the boundary distance effect disappears when the 2D spatial random effect term is included in the model. The expectation was that it might not survive in some cases. In order to address that research question we needed to identify boundary regions where the effect is most pronounced. Any demonstration that the effect does not survive would have little relevance had we examined regions where the effect is small at the outset. For this reason we performed a visual inspection of various images together with boundary ADC plots to identify boundary regions where the effect is most pronounced. We believe that this was a logical starting point and the correct strategy, given the primary objective.Having established, in each of the 5 cases that were examined, that the boundary effect was not obliterated by the addition of the 2D random effect term, the research objectives would change moving forward towards the ultimate clinical question. The main motivation for this study is existing literature indicating that the ADC-gradient at the tumour boundary provides a useful prognostic indicator. We have cited a number of key papers on survival models based on dMRI measures. Some researchers have generated an indicator variable based on a boundary-averaged measure. But a number of questions must be addressed. Tumour development and ultimate patient survival might not be strongly related to average boundary characteristics as opposed to local boundary structure. Obviously, any attempt to obtain a good understanding of the relationship between tumour structure at the boundary, using dMRI as a probe, and patient survival is a huge undertaking. A reliable tumour segmentation method is an absolute requirement. In addition, the current model would require modification to incorporate a position dependence into the coefficient that captures the gradient in ADC at the boundary. (Dr. Porter is correct in his assertion that the magnitude of this coefficient, as obtained for a selected ROI, will depend on position.) The exact form of the position-dependent boundary distance term requires investigation. Although averaging over the boundary offers a simple solution, we assume that important structural information would be lost if an averaged coefficient was adopted as a prognostic summary measure.So, moving forward, it would be wrong to continue with the manual boundary selection method adopted in the present study. Elucidation of the relationship between boundary structure, tumour development and patient survival will be a huge undertaking. But, as stated above, the objective underlying the current study was much less ambitious than a comprehensive characterisation of the entire tumour boundary. Instead, we undertook an examination of the relative magnitude of the random 2D spatial variation in ADC, compared with the boundary distance effect, in a region where the latter was most pronounced. Given this objective, it was logical to select a boundary ROI where the boundary effect is particularly marked, based on visual inspection.Comment 4. I am not convinced by the use of the ADC0.5 parameters, whose definition seems quite contrived and not very intuitive.Response to Comment 4. We thank Dr. Porter for suggesting that we examine the delta parameter as an alternative to ADC0.5. In an attempt to reduce the length of the original paper, we did not offer much explanation for focussing on the latter. We wanted to identify a univariate signal-attenuation measure that should be sensitive to a range of component diffusivities, and thus address the problem that ADCs fails to capture the signal attenuation that occurs at low b-values. We suggest that the ADC0.5 summary measure offers a number of advantages, including the fact that it has the dimensions of a diffusion coefficient. (We use the term 'summary measure' in the sense of Matthews et al., 1990, BMJ, 300, 230-235, where a univariate statistic is used to capture a key feature of the complex dependence observed in a set of repeated measurements, repeated with differing b-values in the present context.)To expand on our reason for using ADC0.5, we remind readers that the main motivation for our study was existing literature indicating that tumour boundary heterogeneity carries prognostic information. Ultimately, some prognostic-indicator selection criterion might be adopted as a mechanism for deciding which summary measures are most useful. If ADC0.5 is found to be lacking as a prognostic indicator, then it should be abandoned, but we cannot know, in advance, which measures will be most useful. It may, indeed, turn out that the boundary distance effect, as characterised by delta, is better than ADC0.5 as a prognostic indicator.There are a number of reasons why we did not focus on delta in our original analysis. Firstly, it does not have the dimensions of a diffusion coefficient. In contrast, ADC0.5 offers this advantage, as does any other summary measure designed to capture signal attenuation over a defined b-value range. Some b-value less than 500 s/mm**2 could be selected to generate an ADC summary measure that is dominated by the faster diffusion components. In the absence of a comprehensive set of observations over a range of b-values we have no information on the shape of the attenuation curve at low b-value, so the value used to derive an ADCfast summary measure remains arbitrary. The point is, however, that given the MCMC output obtained from these analysis, it is possible to calculate an ADCfast summary measure, analogous to ADC0.5, but designed to capture the rapidly attenuated part of the decay curve. (The approach is reminiscent of curve peeling.) The distinction is between using the b-value at half-maximum intensity to calculate ADC0.5, and thus obtain a univariate summary measure designed to capture both the fast and slow diffusion components, versus using some ADC score, designed to capture the fast diffusion components.Some readers might dismiss this approach as arbitrary. As a rejoinder we would point out that, given the structural complexity of all biological tissue, any diffusion coefficient derived from DW images must be regarded as empirical (hence the term 'apparent diffusion coefficient'). The fact is that diffusion barrier structures (soluble and insoluble) exist over a huge range of scales, relative to the free diffusion path length. In this sense all CNS tissue ADC estimates are empirical. As stated previously, the important point is whether the selected measure carries prognostic information. ADC0.5, delta or any other empirical parameter derived from the DWI data might be examined from this perspective.We wish to clarify another point related to Dr. Porter's comment on using delta as a summary measure, in preference to ADC0.5. As stated above, we regard ADC0.5 as a univariate summary measure, designed to capture signal attenuation attributable to a range of diffusivities. The question is whether this is desirable, as opposed to using a bivariate measure based on ADCs, combined with some ADCfast parameter. Thinking again in terms of prognostic modelling as the ultimate objective, a univariate score does offer a simplicity that will be lost if one adopted a bivariate measure. We have no preconceived opinion regarding which boundary effect measure will turn out to be most useful as a prognostic indicator. But we suggest that ADC0.5 should be retained as a candidate because it offers both simplicity and has the dimensions of a diffusion coefficient.Regarding the latter point, it might be noted that it is not uncommon to see biomedical univariate composite scores derived by combining measures having unrelated dimensions. The resulting composite scores have no defined dimension, and the weighting assigned to the component measures is arbitrary. Although a composite score might be generated by combining the delta and ADCs estimates, in our view this approach is not appealing. The ADC0.5 summary measure does not fall into this composite score category.A final point regarding the alternative bivariate/multivariate approach to characterising the multi-exponential decay curve, in favour of using a univariate summary measure. We do not reject the former, recognising that the delta-ADCs pair of parameters might be chosen for this purpose. But, a point in favour of adopting a univariate summary measure applies to studies involving frequentist hypothesis testing. It is an unfortunate fact that a complication arises in multivariate analyses. A number of standard multivariate tests have been devised, but these can yield apparently contradictory results, depending on the nature of the departure from the null (see, for example, Olson, 1976, On choosing a test statistic in multivariate analysis of variance, Psychological Bulletin, 83, 579-586). This problem does not arise in the univariate case. But we do acknowledge the following: to suggest that a univariate summary measure (e.g., ADC0.5) avoids this complication merely hides the underlying non-uniqueness problem (the chosen summary measure will not be unique).To conclude this response to Dr. Porter's comment on using ADC0.5 as a summary measure, we should emphasize that we do not dismiss the suggestion that delta might provide an alternative characterisation of tumour heterogeneity and that this parameter might be added to the list of potential prognostic indicators. Our current paper does downplay any primary role for delta in this context (we thank Dr. Porter for alerting us to this) and the Results section could be extended to include the relevant statistics. That said, we trust that our argument for using ADC0.5 as a univariate summary measure addresses the reservations that Dr. Porter has expressed on this matter.Comment 5. As a point of reference, it would be interesting to see the corresponding results for spatial heterogeneity in regions of brain tissue with ADC values in the normal range or for brains of healthy volunteers. This might help to establish to what extent the large subject-dependent variation in spatial heterogeneity in the clinical data is meaningful.Response to Comment 5. Among the reasons for restricting our analyses to pathological tissue is that the main focus of the study is an examination of the tumour boundary-distance effect. Given the length of the current paper we did not consider the possibility of including results obtained from normal tissue. But we do agree entirely with this comment. A within-subject comparison of the tumour core with contralateral tissue would be a manageable undertaking. Given the marked differences between individual tumours, a meaningful group comparison with a group of normal subjects would be more challenging.Comment 6, page 13, Discussion: The authors state that the proposed MCMC method is expected to produce ADC range data that is less extreme than values provided by an independent voxels analysis. To substantiate this claim, it would have been helpful to provide a direct comparison of results generated by applying both methods to the current data set.Response to Comment 6. Random effects shrinkage is very well established (see Chapter 5 in ref 23, for example). The main reason for not showing results obtained using simple regression analyses based on the iid (identically and independently distributed) residuals assumption is the invalidity of this assumption given spatially correlated data. Random effects modelling is an established approach to dealing with correlated (clustered) data where the independent observations assumption does not apply. Restated, in order to examine the water diffusivity dependence on distance from the tumour boundary, we used a random effects model to circumvent the invalidity of a regression analysis based on the independent-observations assumption. That said, we can compare the voxel-specific ADC estimates obtained from an independent-voxels analysis and those given by our random effects model. This provides a mechanism for showing the shrinkage that often occurs among the estimates given by a random effects model.In order to provide an example of the shrinkage obtained in a typical random effects analysis, this link provides a figure showing the ADCs estimates obtained for the tumour core region in one of the five subjects (using core data to avoid the boundary effect complication). The plot shows the ADC values obtained using an independent voxels, MCMC 2-point (b500, b1000 observations) analysis compared with the estimates given by the spatial model outlined in the paper. Obviously the difference between the two treatments extends beyond the random effect terms included in the Bayesian spatial model. Nevertheless the plots serve to show the considerable shrinkage often observed under a random effects model, given sparse and noisy observations. It might be noted that a typical 2-stage summary measures assessment of DW-image spatial heterogeneity would be based on a first stage voxel-by-voxel (independent voxels) fitting of the signal intensity data (a procedure similar to that used to produce the estimates in the LHS column of the figure), followed by a second-stage assessment of dispersion among the resulting ADC estimates.Comment 7. The field strength of scanner was omitted from the Methods(1.5T).Response to Comment 7. The next revision should be modified to indicate a field strength of 1.5T.Comment 8. Please state in the Methods what type of diffusion preparation was used (Stejskal-Tanner or twice-refocused). I assume that the twice-refocused method was used ...Response to Comment 8. The next revision should be modified to indicate that the twice refocused preparation method was adopted.Comment 9. The symbols used for signal offset at b=0 (equations 3,5,6) and the set of nearest-neighbour voxel indices (equations 12,13) are superficially similar and this might cause confusion.Response to Comment 9. To avoid confusion the next revision should be modified by adding the following to the sentence that includes equations (12) and (13):(the lower-case del (partial) symbol used in equations 12 and 13 should not be confused with delta, which appears in equations 3 et seq.)"
}
]
}
] | 1
|
https://f1000research.com/articles/5-2082
|
https://f1000research.com/articles/5-753/v1
|
26 Apr 16
|
{
"type": "Observation Article",
"title": "Respondent driven sampling of wheelchair users: A lack of traction?",
"authors": [
"John A. Bourke",
"Philip J. Schluter",
"E. Jean C. Hay-Smith",
"Deborah L. Snell",
"Philip J. Schluter",
"E. Jean C. Hay-Smith",
"Deborah L. Snell"
],
"abstract": "Background: Internationally wheelchair users are an emerging demographic phenomenon due to their rapidly increasing life-span coupled with accelerated general population ageing. While having significant healthcare and social implications, basic robust epidemiological information of wheelchair users is often lacking due in part to this population’s “hidden” nature. Increasingly popular in epidemiological research, Respondent Driven Sampling (RDS) provides a mechanism for generating unbiased population-based estimates for hard-to-reach populations, overcoming biases inherent within other sampling methods. This paper reports the first published study to employ RDS amongst wheelchair users.Methods: Between October 2015 and January 2016, a short, successfully piloted, internet-based national survey was initiated. Twenty seeds from diverse organisations were invited to complete the survey then circulate it to peers within their networks following a well-defined protocol. A predetermined reminder protocol was triggered when seeds or their peers failed to respond. All participants were entered into a draw for an iPad. Results: Overall, 19 people participated (9 women); 12 initial seeds, followed by seven second-wave participants arising from four seeds. Completion time for the survey ranged between 7 and 36 minutes. Despite repeated reminders, no further people were recruited. Discussion: While New Zealand wheelchair user numbers are unknown, an estimated 14% of people have physical impairments that limited mobility. The 19 respondents generated from adopting the RDS methodology here thus represents a negligible fraction of wheelchair users in New Zealand, and an insufficient number to ensure equilibrium. While successful in other hard-to-reach populations, applying RDS methodology to wheelchairs users requires further consideration. Formative research exploring areas of network characteristics, acceptability of RDS, appropriate incentive options, and seed selection amongst wheelchair users is needed.",
"keywords": [
"Wheelchair users",
"Disability",
"Respondent driven sampling",
"Social epidemiology",
"Sampling approaches"
],
"content": "Introduction\n\nRobust epidemiological research generally requires data collection from representative samples of the population of interest, and effective modes of sampling contact are essential1. Such effective modes can be difficult in hard-to-reach populations where no (or inadequate) sampling frames exist. Traditional chain-referral sampling approaches are inherently biased in their participant selection methods; a bias that is compounded as recruitment waves continue. Respondent Driven Sampling (RDS) was developed to counter these biases, employing specific data collection and statistical analysis methods which enable valid population-based estimates2–4. Despite wider adoption of RDS, and its successful application in many topic areas, methodological concerns have been raised. RDS estimates are, at times, more variable than expected5, and some sampling patterns appear to violate core RDS assumptions6,7.\n\nThe prevalence of wheelchair users has rapidly increased over the last half century due in part to advancing medical care, ageing populations, increasing community supports, increased prescription of wheelchairs, and changes in attitudes to disablement such that people may feel less stigmatised about using a wheelchair8,9. Despite this, robust epidemiological research in this group is scant10. Contacting wheelchair users in the community is challenging. Recruitment approaches are often limited to using disability organisations and personal contacts, which can differentially exclude many wheelchair users11. Consequently, wheelchair users frequently constitute a ‘hidden population’, under-researched and excluded from population estimates12. Furthermore, many countries, including New Zealand, have yet to establish registries of wheelchair users which could provide a sampling frame13.\n\nHere we report our experience of applying a RDS methodology to a survey of wheelchair users in New Zealand. To our knowledge this is the first time RDS has been applied to people who use wheelchairs, and could potentially offer a significant new sampling approach in epidemiology and disability fields.\n\n\nMethods\n\nOpen from October 2015 until January 2016, this study employed a short internet-based national survey. Administered through the Surveymonkey™ website, an information sheet and video were embedded within the survey preamble (see Supplementary material). The information sheet stated that informed consent was implied through the voluntary participation in the survey. Ethical approval was obtained from the University of Canterbury Human Ethics Committee (reference HEC 2015/117). Eligibility criteria included: wheelchair use as the primary form of mobility; being a New Zealand resident; aged 16 years or more; being able to read English; having internet access; and, having an operational email account.\n\nInvitations seeking ‘seed’ participants were circulated to various national disability organisations serving members with a range of impairments that lead to wheelchair use, including the New Zealand Spinal Trust, the Earthquake Disability Leadership Group, the Multiple Sclerosis Society of New Zealand, the Cerebral Palsy Society of New Zealand, and CCS Disability Action. People expressing interest in being seeds contacted the researcher, who confirmed eligibility and then sent a recruitment code and a link to the survey website. Once a participant completed the survey, they were thanked and emailed three unique recruitment codes. Participants were asked to email one code and the survey link to three other persons they knew who were likely to satisfy the eligibility criteria. This process was envisaged to continue for multiple recruitment waves. Participation was incentivised (an entry into a draw to win an iPad); one entry for completing the survey, and another when each person they recruited completed the survey. Recruitment chains were tracked through tracing the recruitment codes. A predetermined reminder protocol was triggered when seeds or their peers failed to respond.\n\n\nResults\n\nTwelve seed participants completed the survey. A further seven participants were recruited from the second wave, from a total of four seeds, and none were recruited from the third wave. The final sample size (n=19) failed to satisfy the statistical requirements needed to reach equilibrium, the point at which the sample composition becomes independent of the initial seeds, thereby enabling the calculation of unbiased population estimates4. Mean age of participants was 55.6 years (range: 28–73 years), and nine were women. Eighteen were identified as New Zealand European and one was identified as New Zealand Māori. Reasons for using a wheelchair included impaired mobility resulting from spinal cord injury (n=10), cerebral palsy (n=2), spina bifida (n=2), muscular dystrophy (n=2), poliomyelitis (n=2), and arthritis (n=1). Survey completion time ranged between 7 and 36 minutes.\n\n\nDiscussion\n\nDespite a rigorous recruitment process and offering incentivising participation, our use of RDS failed as an effective sampling approach amongst wheelchair users in New Zealand. There are a number of possible explanations as to why this occurred. The target population of the study was novel compared with hidden populations generally targeted by RDS studies. Research using RDS typically samples stigmatised populations, such as those with greater risk of HIV, including injecting drug users6, men who have sex with men7,14, and sex workers15. Wheelchair users have experienced increased integration into many societies in recent years and are arguably less stigmatised when compared to populations traditionally sampled using RDS. Although the precise mechanism by which perceived stigma might affect RDS participation is unknown it, nonetheless, remains noteworthy. Second, the use of an unguaranteed reward (entry into a draw for an iPad) for survey completion has not been previously reported in RDS studies. This lack of guaranteed reward may have influenced participation. In addition, RDS studies often offer participants additional non-monetary free services related to the mitigation of HIV risk, such as counselling and educational material16.\n\nUntil such time as these factors, and their implications for recruitment, are better understood we feel that using RDS for recruiting wheelchairs users may have limited merit, and recommend formative research to optimise success. Exploring the areas of network characteristics, acceptability of RDS, appropriate incentive options, and seed selection have all been suggested as important for assessing the feasibility and appropriateness of RDS in certain populations15. Specifically, formative research regarding specific seed selection is warranted. Motivated seeds with large network contacts can improve recruitment effectiveness. Indeed, one RDS study exploring people who inject drugs in Sydney Australia reported that 80% of their participants resulted from one seed17.\n\n\nConclusions\n\nWheelchair users are an increasingly prevalent population in society who often lack an adequate sampling frame, and sampling approaches enabling valid population based estimates are becoming increasingly necessary. This paper reported the failure of RDS to survey wheelchair users. Despite the unsuccessful recruitment in this study, further research exploring the application of RDS with wheelchair users is recommended before discounting this sampling approach in this population.\n\n\nData availability\n\nData are available upon request from the corresponding author to protect participant identity. Demographic data will be pooled to protect participant identity, as individual-level demographic data could be theoretically traceable due to the small sample size, and suspected small national population of wheelchair users.\n\n\nConsent\n\nAll participants were informed that the voluntary completion of the survey implied informed consent, including for the publication of survey data.",
"appendix": "Author contributions\n\n\n\nAll authors contributed to the design of the study. JAB conducted the data collection and prepared the first draft of the manuscript. All authors contributed to subsequent drafts and the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis paper reports on a section of JAB’s doctoral research, which was supported by a University of Canterbury Doctoral Scholarship and a research scholarship from the Burwood Academy of Independent Living.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nInternet-based national survey.\n\nClick here to access the data.\n\nInformation sheet.\n\nClick here to access the data.\n\n\nReferences\n\nSinclair M, O’Toole J, Malawaraarachchi M, et al.: Comparison of response rates and cost-effectiveness for a community-based survey: postal, internet and telephone modes with generic or personalised recruitment approaches. BMC Med Res Methodol. 2012; 12: 132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeckathorn DD: Respondent-driven sampling: A new approach to the study of hidden populations. Soc Probl. 1997; 44(2): 174–199. Publisher Full Text\n\nWhite RG, Hakim AJ, Salganik MJ, et al.: Strengthening the Reporting of Observational Studies in Epidemiology for respondent-driven sampling studies: “STROBE-RDS” statement. J Clin Epidemiol. 2015; 68(12): 1463–1471. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVolz E, Heckathorn DD: Probability based estimation theory for respondent-driven sampling. J Off Stat. 2008; 24(1): 79–97. Reference Source\n\nGoel S, Salganik MJ: Assessing respondent-driven sampling. Proc Natl Acad Sci U S A. 2010; 107(15): 6743–6747. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYoung AM, Rudolph AE, Quillen D, et al.: Spatial, temporal and relational patterns in respondent-driven sampling: evidence from a social network study of rural drug users. J Epidemiol Community Health. 2014; 68(8): 792–798. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhillips G 2nd, Kuhns LM, Garofalo R, et al.: Do recruitment patterns of young men who have sex with men (YMSM) recruited through respondent-driven sampling (RDS) violate assumptions? J Epidemiol Community Health. 2014; 68(12): 1207–1212. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSapey B, Stewart J, Donaldson G: Increases in wheelchair use and perceptions of disablement. Disability & Society. 2005; 20(5): 489–505. Publisher Full Text\n\nRussell JN, Hendershot GE, LeClere F, et al.: Trends and differential use of assistive technology devices: United States, 1994. Adv Data. 1997; (292): 1–9. PubMed Abstract\n\nThe New Zealand Convention Coalition: The Second Report of the Independent Monitoring Mechanism on the Convention of the Rights of Persons with Disabilities: Making Disability Rights Real Whakatūturu ngā Tika Hauatanga. Wellington: The New Zealand Convention Coalition, 2014. Reference Source\n\nEdwards K, McCluskey A: A survey of adult power wheelchair and scooter users. Disabil Rehabil Assist Technol. 2010; 5(6): 411–419. PubMed Abstract | Publisher Full Text\n\nSmaill RP: Ageing with Spinal Cord Injury in New Zealand. Christchurch: University of Otago, 2014. Reference Source\n\nFitzgerald SG, Kelleher A, Teodorski E, et al.: The development of a nationwide registry of wheelchair users. Disabil Rehabil Assist Technol. 2007; 2(6): 358–365. PubMed Abstract | Publisher Full Text\n\nStrömdahl S, Lu X, Bengtsson L, et al.: Implementation of Web-Based Respondent Driven Sampling among Men Who Have Sex with Men in Sweden. PLoS One. 2015; 10(10): e0138599. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimic M, Johnston LG, Platt L, et al.: Exploring barriers to 'respondent driven sampling' in sex worker and drug-injecting sex worker populations in Eastern Europe. J Urban Health. 2006; 83(6 Suppl): i6–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalekinejad M, Johnston LG, Kendall C, et al.: Using respondent-driven sampling methodology for HIV biological and behavioral surveillance in international settings: a systematic review. AIDS Behav. 2008; 12(4 Suppl): S105–S130. PubMed Abstract | Publisher Full Text\n\nPaquette DM, Bryant J, Crawford S, et al.: Conducting a respondent-driven sampling survey with the use of existing resources in Sydney, Australia. Drug Alcohol Depend. 2011; 116(1–3): 125–131. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "13572",
"date": "09 May 2016",
"name": "Jesse Kokaua",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI think this paper has merit as a scientific publication, in that it adds to the body of research about RDS by its application to a \"hidden\" population, that by all accounts should be reasonable, but turned out to be not effective.The abstract begins with a sentence that ends with a repeated point \"rapid increasing life-span\" and \"accelerated general population ageing\". One of those statements is redundant. Paragraph 2 of the introduction starts with international findings but does the second sentence beginning with \"Despite this . . \" refer to the NZ experience, or outside of the US and UK?I feel that some points could be expanded upon:The authors could expand, in the introduction, how the method improves the validity of, and reduces biases in, population-based estimates over other methods.Examples of typical target populations for this method that are introduced in the discussion should move to the introduction, as it supports the novelty of this application.I feel the results are two-fold. Firstly, the sample as described. I know its a moot point but I would be interested to know how big a sample is required to reach equilibrium. Secondly, since this is a paper on the RDS, I would have liked to see a description of the total numbers of people approached in each wave, with eligible and ineligible participants. Thus reporting if the failure was a result of response, eligibility, or simply a lack of invitation to participate in subsequent waves.This is an interesting application of a novel method of sampling from a group of people made hard-to-reach, it seems, as a result of their successful integration into contemporary New Zealand society. One characteristic of strongly stigmatised groups in New Zealand, such as those identified in the paper, are their tendency to establish strong internal networks, both informal and formal, that I feel would be useful to this type of methodology. It would be difficult for individuals with small networks to fulfill the random selection criteria for RDS. I take it that this may have influenced the less stigmatised wheelchair user population but it wasn't clearly stated if this was the case.The paper reports a good example of where a useful method may not apply, in its initial format, to a particular interest group.However, it wasn't totally discounted as a potential approach but the reader is left without any suggestion as to how it might be modified to improve its take up, except for the suggestion of formative research, while fulfilling the assumptions required for RDS to provide unbiased population-wide estimates.",
"responses": [
{
"c_id": "2155",
"date": "26 Aug 2016",
"name": "Philip Schluter",
"role": "Author Response",
"response": "Thank you greatly for your prompt, encouraging, and constructive feedback on our paper. Together with the comments made by the other two reviewers, your feedback has resulted in what we feel is a much improved paper. We would like to take this opportunity to make two notes. [1] The original submission was restricted to 1000 words in total – but, with permission from the journal, this constraint has been relaxed and we are able to include more material. This means we are able to broaden discussion of relevant issues and thus strengthen the manuscript. [2] This paper was submitted to the journal as an ‘observation article’, defined as: Observation Articles allow the description of a novel observation that may be unexpected, and possibly currently without explanation. An observation can be a phenomenon that has been identified in field work, in the laboratory or through experimental analysis (see: http://f1000research.com/for-authors/article-guidelines/observation-articles). I think this paper has merit as a scientific publication, in that it adds to the body of research about RDS by its application to a \"hidden\" population that by all accounts should be reasonable, but turned out to be not effective. Thank you for your positive comments and summation. This, for us, was the basis for opting for the “novel observation” article type. The abstract begins with a sentence that ends with a repeated point \"rapid increasing life-span\" and \"accelerated general population ageing\". One of those statements is redundant. Although the intention was to make two important points, the first referring to wheelchair users and the second referring to the general population, we agree that this is unnecessarily confusing. We have revised this sentence to: “Internationally, wheelchair users are an emerging demographic phenomenon, due to their rapidly increasing life-span.” Paragraph 2 of the introduction starts with international findings but does the second sentence beginning with \"Despite this . . \" refer to the NZ experience, or outside of the US and UK? This appears to be an international trend. However, to be clearer we have included “in New Zealand and Australia” and inserted an apposite reference. I feel that some points could be expanded upon: The authors could expand, in the introduction, how the method improves the validity of, and reduces biases in, population-based estimates over other methods. Agreed – indeed, a similar point was raised by another reviewer. Our original submission was limited in scope due to the word restrictions, as outlined above. However, in this revision, we have included more information and discussion. Specifically, we have included text providing a brief introduction to RDS (See second paragraph in Introduction section). Examples of typical target populations for this method that are introduced in the discussion should move to the introduction, as it supports the novelty of this application. Thank you for this comment; it indeed strengthens our introduction. In our revised manuscript we have included the following sentence with paragraph one of the Introduction: “RDS has traditionally been used to sample ‘hidden’ populations with inadequate sampling frames, such as those with greater risk of HIV, including injecting drug users.” I feel the results are two-fold. Firstly, the sample as described. I know it’s a moot point but I would be interested to know how big a sample is required to reach equilibrium. We also felt that this information would strengthen our paper, but omitted it initially – due to the word count restrictions. In this version, within the Introduction, we now explicitly add an explanation about the concept of equilibrium and how it is reached (see third paragraph), while the second paragraph in the Results section explains specifically why our sample failed to satisfy the threshold for in-equilibrium data. Secondly, since this is a paper on the RDS, I would have liked to see a description of the total numbers of people approached in each wave, with eligible and ineligible participants. Thus reporting if the failure was a result of response, eligibility, or simply a lack of invitation to participate in subsequent waves. These details would be hard, if not ethically impossible, to determine. Consistent with RDS protocol, an email was sent to people asking them to send on the invitation to three people, but there was no obligation for us (the researchers) to receive confirmation about whether all three recruitment codes were successfully allocated and distributed. Nevertheless, more information has been included in the Methods section (second paragraph) regarding how participants were asked to recruit three others, and the first paragraph in the Results section, which clarifies recruitment and response rates. This is an interesting application of a novel method of sampling from a group of people made hard-to-reach, it seems, as a result of their successful integration into contemporary New Zealand society. One characteristic of strongly stigmatised groups in New Zealand, such as those identified in the paper, are their tendency to establish strong internal networks, both informal and formal, that I feel would be useful to this type of methodology. It would be difficult for individuals with small networks to fulfil the random selection criteria for RDS. I take it that this may have influenced the less stigmatised wheelchair user population but it wasn't clearly stated if this was the case. Yes, this is valid and interesting point and was also raised by another reviewer. In response, we have included two sentences in the first paragraph of the Discussion section addressing the area of perceived stigma and whether or not wheelchair users have sufficiently strong internal networks that enable the random recruitment of other wheelchair users. The paper reports a good example of where a useful method may not apply, in its initial format, to a particular interest group. However, it wasn't totally discounted as a potential approach but the reader is left without any suggestion as to how it might be modified to improve its take up, except for the suggestion of formative research, while fulfilling the assumptions required for RDS to provide unbiased population-wide estimates. Thank you for this comment that will strengthen our paper. In response we have included more information in the second paragraph of the Discussion section about potential ways to improve similar studies, namely, selecting diverse seeds with larger networks and providing greater information and education to those seeds."
}
]
},
{
"id": "14016",
"date": "13 Jun 2016",
"name": "John F. Smith",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI support indexing this article as it as it appears to be a first-time test of the utility of Respondent Driven Sampling (RDS), a relatively novel sampling procedure, on a new \"hidden\" (hard to easily access) population--wheelchair riders. In this case it was not successful in recruiting sufficient respondents, but reporting that, and possible reasons for that outcome, could be useful for refining usage of the procedure.\n\nI think the title and abstract are clear and provide good insight into the major points of the paper.\nIn addition to the possible reasons for low returns provided by the author I think it would be useful to include consideration of;\nThe typically low returns from internet surveys compared to face to face and pencil and paper surveys (see 1,2,3). It is possible that promoting an internet survey was just insufficient to motivate the \"seeds\", or the second wave recruits, to recruit more members into the network, even with an iPad reward incentive.\n\nElaborating on the \"stigmatization\" issue raised in the discussion. The \"mechanism by which perceived stigma may effect RDS sampling\" rather than being unknown, may well be that other groups e.g., drug users, \"men who have sex with men\"(MSM), sex-workers etc as a result of stigma, may have stronger social networks and obvious gathering spaces and be easier to access via seeds. Seeds with these groups are often \"leaders\", advocates who are \"out\", support group leaders, health service liaison/volunteer workers who move easily within those communities, social gatherings, entertainment or work spaces. There are not the same social presses for wheelchair riders to have these group social connections or common activities.\n\nMethodological point. The article focuses on RDS's value for accessing representative population samples for research seeking normative and epidemiological data and concomitant importance of avoiding/controlling sample selection bias. However, RDS could also be valuable for accessing qualitative data e.g., wheelchair riders' views/experiences with public policies/services etc where as wide a range of responses (variability) should be expected, indeed welcomed, rather than seen as requiring some statistical weighting or control procedure. Other forms of criterion for data \"completeness\" e.g., saturation could be used here. Another use for RDS is just to collect as large a sample as possible in hidden groups regardless of their representativeness. For example, RDS \"seeds\" have been used to recruit as many MSM as possible in \"test and treat\" outreach campaigns for blood testing for HIV and ante-retroviral treatment programs.\n\nOverall a potentially useful addition to methodological literature on sampling procedures.",
"responses": [
{
"c_id": "2154",
"date": "26 Aug 2016",
"name": "Philip Schluter",
"role": "Author Response",
"response": "Thank you greatly for your prompt, encouraging, and constructive feedback on our paper. Together with the comments made by the other two reviewers, your feedback has resulted in what we feel is a much improved paper. We would like to take this opportunity to make two notes. [1] The original submission was restricted to 1000 words in total – but, with permission from the journal, this constraint has been relaxed and we are able to include more material. This means we are able to broaden discussion of relevant issues and thus strengthen the manuscript. [2] This paper was submitted to the journal as an ‘observation article’, defined as: Observation Articles allow the description of a novel observation that may be unexpected, and possibly currently without explanation. An observation can be a phenomenon that has been identified in field work, in the laboratory or through experimental analysis (see: http://f1000research.com/for-authors/article-guidelines/observation-articles). I support indexing this article as it as it appears to be a first-time test of the utility of Respondent Driven Sampling (RDS), a relatively novel sampling procedure, on a new \"hidden\" (hard to easily access) population - wheelchair riders. In this case it was not successful in recruiting sufficient respondents, but reporting that, and possible reasons for that outcome, could be useful for refining usage of the procedure. Thank you for recognising the novelty of this approach, together with your encouraging and supportive comments regarding our paper. I think the title and abstract are clear and provide good insight into the major points of the paper. Thank you. In addition to the possible reasons for low returns provided by the author I think it would be useful to include consideration of; The typically low returns from internet surveys compared to face to face and pencil and paper surveys (see 1,2,3). It is possible that promoting an internet survey was just insufficient to motivate the \"seeds\", or the second wave recruits, to recruit more members into the network, even with an iPad reward incentive. Yes, this is indeed a valid consideration. In response, we have included a sentence the second paragraph in our revised discussion specifically referring to how more informed and enthusiastic seeds could have encouraged greater response rates, acting as a potential mitigation strategy against the impersonal nature of electronic surveys. Elaborating on the \"stigmatization\" issue raised in the discussion. The \"mechanism by which perceived stigma may effect RDS sampling\" rather than being unknown, may well be that other groups e.g., drug users, \"men who have sex with men\"(MSM), sex-workers etc as a result of stigma, may have stronger social networks and obvious gathering spaces and be easier to access via seeds. Seeds with these groups are often \"leaders\", advocates who are \"out\", support group leaders, health service liaison/volunteer workers who move easily within those communities, social gatherings, entertainment or work spaces. There are not the same social presses for wheelchair riders to have these group social connections or common activities. Yes, this is valid and interesting point, and was also raised by another reviewer. In response, we have elaborated on the stigma issue by including new text in the first paragraph in the Discussion section addressing the area of perceived stigma, and whether or not wheelchair users have a sufficiently strong internal network that enable the random recruitment of other wheelchair users. Methodological point. The article focuses on RDS's value for accessing representative population samples for research seeking normative and epidemiological data and concomitant importance of avoiding/controlling sample selection bias. However, RDS could also be valuable for accessing qualitative data e.g., wheelchair riders' views/experiences with public policies/services etc where as wide a range of responses (variability) should be expected, indeed welcomed, rather than seen as requiring some statistical weighting or control procedure. Other forms of criterion for data \"completeness\" e.g., saturation could be used here. Another use for RDS is just to collect as large a sample as possible in hidden groups regardless of their representativeness. For example, RDS \"seeds\" have been used to recruit as many MSM as possible in \"test and treat\" outreach campaigns for blood testing for HIV and ante-retroviral treatment programs. The value of using the RDS in qualitative research is a very interesting concept, and could be raised in the Discussion. However, on balance we feel it may be beyond the scope of our paper here, which was to discuss our experience of unsuccessfully satisfying the requirements needed to produce a theoretically representative sample of the population with no sampling frame. Overall a potentially useful addition to methodological literature on sampling procedures. We appreciate these words and the additional references offered."
}
]
},
{
"id": "14896",
"date": "11 Jul 2016",
"name": "A James O'Malley",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle and Abstract: The title is catchy and appropriate given the content of the article.\n\nArticle content: The article is very well written and is easy to read and follow. The article would be much improved if it included an introduction to respondent drive sampling (RDS) with particular emphasis on the assumptions required for RDS to work (reaching equilibria and otherwise) written in clear terms to a lay audience. Describe what it means for the sampling process to be in equilibrium. Currently, readers do not have the necessary information to reach an informed conclusion as to why RDS did not work well in this instance or to assess whether it is applicable to their own work.\n\nThe rationale for RDS assumes the existence of a social network in which the probability of a tie between a wheelchair user and a randomly selected wheelchair user exceeds the probability of a tie between a wheelchair user and a nonuser wheelchair user. What evidence that this is true? Do wheelchair users have sufficient social and other relationships importantly enough for them to be able to name sufficient other wheelchair users for the next round of sampling. Is it possible that many wheelchair users are isolates (in the network sense) because they are highly functional in everyday life and thus have reduced need for a community of wheelchair users? I don’t know the answers to these questions and still don’t as the needed background information is not provided in the article.\n\nIt would have been helpful to have been told the current state of knowledge about social networks among wheelchair users along with a connection of such knowledge to the assumptions under which RDS can yield unbiased population-level inferences. I would like to see some discussion of how RDS handles scenarios such as isolates in the wheelchair user network. Presumably, the only way such individuals make it into the sample is if they were randomly selected in the first (i.e., seed) wave.\n\nPlease add some general intuition for how RDS yields unbiased population-level estimates. Discuss the role of and derivation of individuals’ sampling probabilities and thus sampling weights for computing population estimates. For example, the work of Thompson (2006a, 2006b, 2003, 2000 – see below – the 2012 text on Sampling and recent arXiv contributions) should be reviewed and used to inform the article.\n\nReferences:\nThompson, Steven K. Targeted random walk designs. Survey Methodology 2006. 32:11-24 Thompson, Steven K. Adaptive web sampling. Biometrics 2006. 62:1224-1234 Chow, Mosuk, Thompson, Steven K. Estimation with link-tracing sampling designs - A Bayesian approach. Survey Methodology 2003. 29:197-205 Thompson, Steven K., Frank, Ove. Model-based estimation with link-tracing sampling designs. Survey Methodology 2000. 26:87-98\n\nData (if applicable): Details about the design/data for the study are lacking. It is stated that 20 wheelchair users were the initial seeds of which 12 agreed to participate (60% response rate). It then appears as though 4 of these 12 participants culled 7 further participants in the second stage of sampling. Question: Did the remaining 8 seeds not generate any wheelchair users (i.e., does the enrolment rate for the second stage equal 7/12) or was the plan to only ask 33% of the initial seeds for links to trace (in which case the enrolment rate is 7/4)? In the third stage, there was no addition to the sample-size. Were the 7 participants who entered at the second stage each asked to name other wheelchair users; is the enrolment rate at the third stage 0/7 or 0/x, where x<7? Was there any limit as to how many wheelchair users a seed to subsequently enrolled person could name? A lot of these questions would be answerable if the design of the RDS was described in the article!\n\nConclusions: The discussion makes a number of good points that have appear sound and logical. The primary point made is that the current study may have failed because the wheelchair user population has different characteristics from populations where RDS has been used successfully. This is a valid point. But shouldn’t this point have been considered when the study was design? Was there any prior information to inform the parameters of the RDS design? Maybe RDS would work quite well for the wheelchair user population as long as the number of initial seeds is much greater than 20. A more positively-framed Discussion might leave readers with a more balanced appreciation for using RDS in their own studies of wheelchair users or other new populations.",
"responses": [
{
"c_id": "2153",
"date": "26 Aug 2016",
"name": "Philip Schluter",
"role": "Author Response",
"response": "Thank you greatly for your prompt, encouraging, and constructive feedback on our paper. Together with the comments made by the other two reviewers, your feedback has resulted in what we feel is a much improved paper. We would like to take this opportunity to make two notes. [1] The original submission was restricted to 1000 words in total – but, with permission from the journal, this constraint has been relaxed and we are able to include more material. This means we are able to broaden discussion of relevant issues and thus strengthen the manuscript. [2] This paper was submitted to the journal as an ‘observation article’, defined as: Observation Articles allow the description of a novel observation that may be unexpected, and possibly currently without explanation. An observation can be a phenomenon that has been identified in field work, in the laboratory or through experimental analysis (see: http://f1000research.com/for-authors/article-guidelines/observation-articles). Title and Abstract: The title is catchy and appropriate given the content of the article. Thank you for the positive comment. Article content: The article is very well written and is easy to read and follow. The article would be much improved if it included an introduction to respondent drive sampling (RDS) with particular emphasis on the assumptions required for RDS to work (reaching equilibria and otherwise) written in clear terms to a lay audience. Describe what it means for the sampling process to be in equilibrium. Currently, readers do not have the necessary information to reach an informed conclusion as to why RDS did not work well in this instance or to assess whether it is applicable to their own work. We thank the reviewer for describing our article as well written and easy to read and follow. We agree that an introduction to RDS would improve the article. Our original submission was limited by word restrictions and article scope, as outlined above. However, in this revision, we have included more details. Specifically, we have included text in the Introduction section (see second paragraph) providing a brief introduction to RDS, and describing equilibrium (see third paragraph). We have expanded the Results section (second paragraph) to describe why equilibrium was not reached in our study. The rationale for RDS assumes the existence of a social network in which the probability of a tie between a wheelchair user and a randomly selected wheelchair user exceeds the probability of a tie between a wheelchair user and a nonuser wheelchair user. What evidence that this is true? We had no evidence to suggest this was true or false and felt on balance that the use of RDS in our study was warranted as there was a reasonable enough chance that it could be true without evidence to the contrary. Conducting a revised search on PubMed, Google scholar, and Science Direct, we could find no literature explicitly addressing this issue (we have included a sentence in the first paragraph of the Discussion section describing this). Do wheelchair users have sufficient social and other relationships importantly enough for them to be able to name sufficient other wheelchair users for the next round of sampling. Is it possible that many wheelchair users are isolates (in the network sense) because they are highly functional in everyday life and thus have reduced need for a community of wheelchair users? I don’t know the answers to these questions and still don’t as the needed background information is not provided in the article. This is indeed possible. The fact that a person uses a wheelchair does by no means infer that they know other people who use wheelchairs, let alone have strong internal networks with other wheelchair users. With respect to these questions we have included new text in the Discussion section (see first paragraph) regarding stigma and whether or not wheelchair users have sufficiently strong internal networks that enable random recruitment of other wheelchair users. It would have been helpful to have been told the current state of knowledge about social networks among wheelchair users along with a connection of such knowledge to the assumptions under which RDS can yield unbiased population-level inferences. I would like to see some discussion of how RDS handles scenarios such as isolates in the wheelchair user network. Presumably, the only way such individuals make it into the sample is if they were randomly selected in the first (i.e., seed) wave. As noted above, the current state of knowledge regarding the social networks among wheelchair users is unknown. As a result we cannot know if some wheelchair users are isolates. RDS theory assumes that the population being sampled comprises of a complete social network component, so in theory every person within their population has a probability of being sampled. In practice, to overcome isolated subpopulations, it is recommended that seeds from diverse subpopulations be selected. We have included a sentence to this effect in the second paragraph of the Discussion section. Please add some general intuition for how RDS yields unbiased population-level estimates. Discuss the role of and derivation of individuals’ sampling probabilities and thus sampling weights for computing population estimates. Due to original word count, we had not included information clarifying that question two of our survey asked participants to provide an estimate of their network size (a requirement of RDS). Had our study produced in-equilibrium data (see new text in second paragraph in Introduction section, and new text in second paragraph of the Results section), sampling weights would have been allocated accordingly with those with smaller social network sizes being given a higher weight. For example, the work of Thompson (2006a, 2006b, 2003, 2000 – see below – the 2012 text on Sampling and recent arXiv contributions) should be reviewed and used to inform the article. References: Thompson, Steven K. Targeted random walk designs. Survey Methodology 2006. 32:11-24. Thompson, Steven K. Adaptive web sampling. Biometrics 2006. 62:1224-1234. Chow, Mosuk, Thompson, Steven K. Estimation with link-tracing sampling designs - A Bayesian approach. Survey Methodology 2003. 29:197-205. Thompson, Steven K., Frank, Ove. Model-based estimation with link-tracing sampling designs. Survey Methodology 2000. 26:87-98. Thank you greatly for suggesting these references, which were reviewed and indeed helped with our thinking. Because our study failed to produce in-equilibrium data that warranted further analysis and consideration of sampling weights and sampling probabilities, substantial engagement with the topics in these papers was felt to be beyond the scope of our paper. Data (if applicable): Details about the design/data for the study are lacking. It is stated that 20 wheelchair users were the initial seeds of which 12 agreed to participate (60% response rate). It then appears as though 4 of these 12 participants culled 7 further participants in the second stage of sampling. Question: Did the remaining 8 seeds not generate any wheelchair users (i.e., does the enrolment rate for the second stage equal 7/12) or was the plan to only ask 33% of the initial seeds for links to trace (in which case the enrolment rate is 7/4)? All 12 seeds were asked to recruit 3 peers and for links to trace. Despite this, 8 seeds did not generate any more participants. The enrolment rate was 7/12. To clarify recruitment/response rates, text has been included in the first paragraph of the Results section. In the third stage, there was no addition to the sample-size. Were the 7 participants who entered at the second stage each asked to name other wheelchair users; is the enrolment rate at the third stage 0/7 or 0/x, where x<7? The enrolment rate was 0/7. All 7 participants were asked to recruit 3 peers and for links to trace, but no participants were recruited. To clarify recruitment/response rates, text has been included in the first paragraph of the Results section. Was there any limit as to how many wheelchair users a seed to subsequently enrolled person could name? A lot of these questions would be answerable if the design of the RDS was described in the article! Thank you for raising this issue. All seeds and subsequent participants were asked to recruit a maximum of 3 peers and for links to trace. Heckathorn (1997) recommends that each respondent is limited to 3 ID codes to pass on to others, to ensure that a broad array of participants is recruited, and to prevent the emergence of semi-professional recruiters. We have now included two sentences describing this in the second paragraph in the Methods section. Conclusions: The discussion makes a number of good points that have appear sound and logical. The primary point made is that the current study may have failed because the wheelchair user population has different characteristics from populations where RDS has been used successfully. This is a valid point. But shouldn’t this point have been considered when the study was designed? Yes, a valid point. In retrospect the potential impact of this point could have attracted greater consideration. However, we had no evidence to suggest it would be detrimental to the success of our employment of RDS. We also acknowledged that our application of RDS with wheelchair users was unprecedented and novel, and many outcomes were simply unknown. In the absence of any reliable sampling frame or registry, we believed the potential success amongst this population (on balance) was greater than the risks. Was there any prior information to inform the parameters of the RDS design? Maybe RDS would work quite well for the wheelchair user population as long as the number of initial seeds is much greater than 20. A more positively-framed Discussion might leave readers with a more balanced appreciation for using RDS in their own studies of wheelchair users or other new populations. Since our employment of RDS, a number of points, that may have improved our experience, have been considered. These include selecting more diverse seeds, and meeting with seeds to provide more information to them. To summarise these points, new text has been included our revised manuscript (see second paragraph in Discussion section)."
}
]
}
] | 1
|
https://f1000research.com/articles/5-753
|
https://f1000research.com/articles/5-91/v1
|
21 Jan 16
|
{
"type": "Research Article",
"title": "Are scientific abstracts written in poetic verse an effective representation of the underlying research?",
"authors": [
"Sam Illingworth"
],
"abstract": "The central purpose of science is to explain. However, who is that explanation for, and how is this explanation communicated once it has been deduced? Scientific research is typically communicated via papers in journals, with an abstract presented as a summary of that explanation. However, in many instances they may be written in a manner which is non-communicatory to a lay reader. This study begins to investigate if poetry could be used as an alternative form of communication, by first assessing if poetic verse is an effective form of communication to other scientists. In order to assess this suitability, a survey was conducted in which two different groups of participants were asked questions based on a scientific abstract. One group of participants was given the original scientific abstract, whilst the second group was instead given a poem written about the scientific study. Quantitative analysis found that whilst a scientific audience found a poetic interpretation of a scientific abstract to be no less interesting or inspiring than the original prose, they did find it to be less accessible. However, further qualitative analysis suggested that the poem did a good job in conveying a similar meaning to that presented in the original abstract. The results of this study indicate that whilst for a scientific audience poetry should not replace the prose abstract, it could be used alongside the original format to inspire the reader to find out more about the topic. Further research is needed to investigate the effectiveness of this approach for a general audience.",
"keywords": [
"Science Communication",
"Interdisciplinary",
"Communication",
"Publishing"
],
"content": "Introduction\n\nThe central purpose of science is to explain (Purtill, 1970). However, who is that explanation for, and how is this explanation communicated once it has been deduced? Scientific research is typically communicated via papers in journals, but whilst to an insider (i.e. a scientist in that field) these papers and journals represent an efficient and effective way of communicating research, to an outsider (i.e. a member of the general public) what they represent and report on may not be at all clear (Meadows, 1985), and in many instances they may be written with a lexical density that makes them inaccessible to a lay reader (Halliday & Martin, 2003).\n\nAlmost all journals require the authors to provide a word-limited abstract as part of the submission process, and whilst the specifics of these abstracts will vary from journal to journal, their purpose effectively remains the same, with Johnson (1995, pp. 28) defining them as “a concise representation of a document’s contents to enable the reader to determine its relevance to specific information.” If the central purpose of science were indeed to explain, is the central purpose of an abstract therefore a summary of that explanation? Swales (1990) considers a scientific abstract to be a ‘rite of passage’ for gaining entry into the scientific community, and that in order to do so the writer needs to demonstrate an “increasing mastery of the academic dialect” (Orasan, 2001, pp. 2).\n\nOrasan (2001) also observed that many authors of scientific papers do not consider the abstract to be particularly important, arguing that in many cases it is written as a necessity just before the paper is submitted. Is it therefore the case, that rather than being an effective and economical method of communicating the research, that the abstract instead represents a rushed précis of the research findings, with an even higher lexical density than that of the main body of the text? If so, then are they useful to anyone who might consider themselves, or indeed be considered an outsider? And if non-experts are unable to fully grasp the summary of the explanation, then what hope do they have of being able to fully understand the research and its potential relevance to them?\n\nClimate change research is a subject which has potential relevance on a global scale, however Rudiak-Gould (2014) found that whilst members of the general public receive information about climate change through the first-hand experience of its effects on their environment, it is still absolutely necessary to effectively communicate the science to them as well. This is because as well as the difficulty in objectively observing long-term trends, there are other issues in their day-to-day lives that the general public must concern themselves with as well. In their study Rudiak-Gould (2014) worked with the Marshall Island’s indigenous population, the Majuro people, where more pressing concerns than long-term sea-level rise were short-term anxieties related to e.g. fluctuations in the cost of rice. This argument is relevant in other communities, where issues such as job security, energy prices and mortgage rates might well take precedence over concerns relating to climate change. In other words, it is not simply enough to assume that people will take notice of their changing environments and act upon them; instead there needs to be an intervention in terms of the effective communication of what is happening, the consequences of this, and what can be done in order to mitigate the effects. Effective science communication should “facilitate conversations with the public that recognize, respect, and incorporate differences in knowledge, values, perspectives, and goals.” (Nisbet & Scheufele, 2009, pp. 1767).\n\nThe accurate communication of scientific research is also vital so that the general public are aware of the consensus in terms of scientific understanding, and researchers should not forget the ‘moral dimension’ and sense of responsibility in terms of communicating their findings to others (Tickell, 2002). This is particularly prominent for studies discussing the anthropogenic nature of climate change, with John et al. (2013) finding that over 97% of climate change papers published between 1991 and 2011 agreed that climate change is a human-caused phenomenon. However, this is not always the way that this argument is presented, which is why it is absolutely vital that scientists endeavour to make their research as transparent and accessible as possible. It has also been shown that the polarization over the validity of climate change science is reduced when information content is communicated alongside a consideration for cultural meanings (Kahan et al., 2015). Can scientific articles take into account these cultural values, thereby acting as an effective way of communicating information? Likewise, if journals are to act as an effectual conduit between scientists and the general public, then how can abstracts be constructed so as to appeal to the widest possible audience whilst still conveying useful and meaningful information, and is there a medium that can be exploited in order to ease this transition?\n\nThe former United States Poet Laureate Robert Pinsky writes that (Pinsky, 2009, pp. 46):\n\n“Poetry mediates, on a particular and immensely valuable level between the inner consciousness of the individual reader and the outer world of other people.”\n\nSimilarly, the English romantic-era poet Percy Bysshe Shelley noted that “poetry lifts the veil from the hidden beauty of the world” (Shelley, 1888). Could poetry therefore be the medium with which to help encapsulate more general audiences with research findings? Science and poetry have much in common, both in terms of their use of metaphor and their embodiment of process, and as the American poet Robert Kelly noted in his poem ‘Science’ (Kelly, 2007):\n\nScience explains nothing\n\nbut holds all together\n\nas many things as it can count\n\nscience is a basket\n\nnot a religion he said\n\na cat as big as a cat\n\nthe moon the size of the moon\n\nscience is the same as poetry\n\nonly it uses the wrong words.\n\nCould poetry therefore help scientists to choose their words more carefully, thereby helping them to avoid the academic dialect that so often ostracizes the non-expert? There is in fact a historical precedent for science being written in poetic verse, most famously evidenced by the works of Erasmus Darwin (1798). However, rather than an entire journal article penned in poetic verse, might instead their abstracts be written in this style, and in doing so might they then appeal to a wider audience, be more readily understood, and potentially encourage the reader to investigate the topic further? If poetry is to be used to help better communicate scientific abstracts to the general public however, it is first of all important to establish if this form of communique is still useful to the experts in the field. In other words, if poetic verse were to be accepted as a suitable abstract style, then would this still be accessible and informative to other scientists? It is the purpose of this study to determine if this new format means that the abstract is still a useful commodity to the ‘insiders.’\n\nThis paper is organised as follows: the methodology used in this study is described, followed by a presentation and discussion of the results from this study. Finally, some perspectives on this work are outlined, discussing what the results imply for future work and for the scientific abstract in general.\n\n\nSurvey selection\n\nIn order to assess the suitability of using poetry in scientific abstracts, a survey was conducted in which two different groups of participants were given an abstract relating to a scientific paper, before being asked questions based on this abstract. One of the groups of participants was given the original scientific abstract, whilst the second group were given a poem that was written about the scientific study. Apart from this the two surveys were identical, and the survey was conducted using the free online survey builder ‘Typeform’ (www.typeform.com), comprising of nine questions delivered with a mixed-method approach. Of these nine questions, five of them asked for demographic information, whilst the remaining four were all related to the abstract itself, asking the participants to sum up in their own words what they thought that the research was about, and also asking them to mark the abstract out of ten (zero being the least) for how accessible and interesting they found the abstract, as well as how likely they were to go and find out more about this research as a result of reading the abstract. A copy of the questionnaire can be found in the Supplementary Materials section of this article, and the non-demographic questions are given below (please note that the numbering of these questions is different to how they appeared in the survey):\n\nQ1 How accessible did you find the abstract? (mark from 0 to 10, with 0 being the least)\n\nQ2 How interesting did you find the abstract? (mark from 0 to 10, with 0 being the least)\n\nQ3 As a result of reading the abstract, how likely are you to go and find out more about this research? (mark from 0 to 10, with 0 being the least)\n\nQ4 After reading the abstract, what do you think that this research is about? (open-ended)\n\nThe choice of the abstract and poem themselves was important, as it was necessary to choose an abstract that was well written so as not to bias the results of the study, it was also important to choose a topic that would be of potential relevance to non-experts. As discussed in the introduction, research concerning climate change demands to be communicated, because of its global relevance and the potential societal consequences of its findings. Ideally then, the scientific study in question would be related to climate change, and would have a well-written abstract.\n\nI write a weekly blog (http://thepoetryofscience.scienceblog.com/), in which I communicate recent scientific research to the general public by reading journal articles and then writing a poem that summarises these findings. From the archive of these poems, there was one which was written about a study into the projected deglaciation of western Canada (Clarke et al., 2015). This study related to climate change, was extremely well written, and was published in a very reputable journal. Glaciers represent natural hazards to local communities and beyond because of their importance to regional water resources (Marshall, 2014), as well as the danger that they pose in relation to outburst floods brought about by a warming climate (see e.g. Bolch et al., 2008). The communication of the impact of climate change on glacial retreat is therefore important, not only for those downstream of the glaciers themselves (Vuille et al., 2008), but also for the wider global communities that are affected by the changes to the global carbon budget and ocean circulation that can be brought about my glacial change (Piotrowski et al., 2005). The abstract for the Clarke et al. (2015) paper, as well as the accompanying poem were thus chosen for this study.\n\nGiven that this study aimed to provide an initial insight into the plausibility of using poetic verse in scientific abstracts, a total sampling size of 100 participants (50 for each survey) was chosen. A convenience sampling strategy was adopted, in which the survey was advertised using Twitter, via both multiple tweets from the author’s Twitter account and the re-tweets that also resulted from this. The target audience were people that identified themselves as being scientists or who had a scientific background, which for the sake of this study were taken to be people that had achieved at least an undergraduate degree in science. This study was carried out according to the British Educational Research Association’s (BERA) ethical guidelines for educational research, with all of the data in this study fully anonymised.\n\nThe abstract from the Clarke et al. (2015) study that was given to the ‘prose’ group of the participants is shown below:\n\n“Retreat of mountain glaciers is a significant contributor to sea-level rise and a potential threat to human populations through impacts on water availability and regional hydrology. Like most of Earth’s mountain glaciers, those in western North America are experiencing rapid mass loss. Projections of future large-scale mass change are based on surface mass balance models that are open to criticism, because they ignore or greatly simplify glacier physics. Here we use a high-resolution regional glaciation model, developed by coupling physics-based ice dynamics with a surface mass balance model, to project the fate of glaciers in western Canada. We use twenty-first-century climate scenarios from an ensemble of global climate models in our simulations; the results indicate that by 2100, the volume of glacier ice in western Canada will shrink by 70 ± 10% relative to 2005. According to our simulations, few glaciers will remain in the Interior and Rockies regions, but maritime glaciers, in particular those in northwestern British Columbia, will survive in a diminished state. We project the maximum rate of ice volume loss, corresponding to peak input of deglacial meltwater to streams and rivers, to occur around 2020–2040. Potential implications include impacts on aquatic ecosystems, agriculture, forestry, alpine tourism and water quality.”\n\nWhilst the poem that was distributed to the ‘poetry’ group is as follows:\n\nIn Canada a study found,\n\nHow glaciers melt in the West.\n\nThe shrinkage is beyond profound,\n\nWith seventy per cent at best;\n\nIf we ignore the Earth’s request\n\nThen ninety-five per cent will go.\n\nNew barren lands will not be dressed,\n\nWith climate change too warm for snow,\n\nThe alpine streams and sapphire lakes they too will go.\n\nOnce the responses to the surveys were collected, the quantitative outcomes were assessed, and the qualitative analysis tool NVivo (Version 10.2.2) was used to perform a qualitative thematic analysis. These results are presented and discussed in the following section.\n\n\nResults and discussion\n\nBox plots of marks out of ten for the responses to the three quantitative survey questions (Q1–Q3) are given in Figure 1–Figure 3, whilst Figure 4 shows the number of words that were written by each of the participants in response to the open-ended qualitative survey question (Q4). As can be seen from Figure 1–Figure 3, the abstracts that were written in the traditional prose format seemed to receive higher marks than their poetic equivalents in terms of accessibility, interest and inspiration (i.e. the likelihood of the reader wanting to find out more about the research topic). Figure 4 also suggests that the average number of words that were written in response to what the scientific study was about was also much higher for the prose group. These differences can also be seen from the mean values that are presented in Table 1. However, in order to ascertain that there really is a marked difference in the average responses to the survey questions, it is necessary to carry out a statistical test to ensure that this is the case.\n\nOutliers are represented by white circles.\n\nOutliers are represented by white circles.\n\nOutliers are represented by white circles.\n\nThe first column corresponds to the survey questions, and the second and third columns give the mean and standard deviations for the poem and prose groups, respectively. The fourth column gives the asymptotic significance (2-tailed) p-value for the Mann–Whitney U test.\n\nQuestions 1–3 are based on an ordinal scale from 0 to 10 (where 0 was the lowest response), as such the responses to these questions will not be normally distributed, and it is therefore necessary to use nonparametric statistics, which make no assumptions about the probability distributions of the variables that are being assessed. Regarding the word count associated with Q4, and because the sample size was relatively small, the Shapiro-Wilk test was used to test for normalisation in the data. For the responses to both the poetry and the prose abstracts it was found that the p-value of the Shapiro-Wilk test was less than 0.001, therefore suggesting that the data significantly deviates from a normal distribution. As such, a nonparametric statistical test was also needed to check if the average word count in the response to Q4 was statistically different between the prose and the poetry groups.\n\nStatistical analysis was performed using the Mann-Whitney U-test, where a p-value of less than 0.05 is considered statistically significant, i.e. there is a statistically significant difference between the responses of the two groups. The Mann-Whitney U-test is the nonparametric equivalent of the independent t-test, and was conducted using IBM SPSS (Version 22.0), the results of which are shown in Table 1. As can be seen from Table 1, at the 95% confidence level the prose group found the abstract more accessible than the poetry group. Similarly, the prose group were statistically more likely to write more than the poetry group in their responses to Q4. With regards to the interest and inspiration of the abstracts, there was no statistically significant difference between the two groups at the 95% confidence level, with the p-values being 0.106 and 0.34, respectively.\n\nFrom the results of the surveys, the prose abstract would appear to be more accessible than the poetry version, with both generating similar levels of interest and inspiration. Given that this research aims to see if abstracts in poetic form are still useful to scientists, from the results of the survey it would appear that they are less useful than a well written piece of prose. It is also worth noting that that the average mark for Q3 (i.e. the extent to which having read the abstract, the reader was inclined to go and find out more about the subject) was below 50% for both formats. It is also a little surprising that the readers found the poem to be no more interesting or inspirational than the abstract, but again this might be down to the high quality of the prose, or alternatively a reflection on the quality of the poem!\n\nSo, from a statistical analysis of the survey it would appear that poetry should not be used as an alternative to prose in the presentation of scientific abstracts, as scientists find this approach to be less accessible. However, before making any further analysis, it is first necessary to carry out a qualitative assessment of the responses to Q4. Could it be that despite ranking the poetic form as being less accessible than the standard format, the participants were still able to successfully deduce the main focus of the research?\n\nFrom the responses to Q4, the qualitative analysis tool NVivo was used to perform a qualitative thematic analysis. An open coding approach was adopted, in which a number of major categories were selected based on the participant’s responses. The responses were then re-examined in order to confirm that the major categories were an accurate representation of the responses. This methodology was adopted for responses from both the poetry and the prose groups, and was carried out until there were no further themes found to be emerging, i.e. until descriptive saturation was reached. Only categories which received a total of more than five responses were considered for analysis. The different major themes, along with the corresponding coding frequencies, are shown in Table 2, whilst Table 3 gives a more detailed description of each of the categories.\n\nThe average number of categories contained within each response for the poetry and prose groups were 1.8 ± 0.97 and 2.5 ± 1.33, respectively. The p-value of the Shapiro-Wilk test for this data was 0.001, therefore suggesting that it significantly deviates from a normal distribution. As such, the Mann-Whitney U-test was used to compare the means, with an asymptotic 2-tailed p-value of 0.011 indicating that at the 95% confidence interval the responses from the prose group had statistically more category groupings per response than those from the poetry group. This result correlates well with the increased word count in Q4 for the prose group, thereby indicating that this group provided more verbose and detailed summaries than the poetry group.\n\nIn terms of the categories themselves, the only category that was applicable to only one group was the ‘Model’ category. This is not surprising, as the poem itself makes no mention of the fact that the study in question was based largely around a set of modelling simulations. This is perhaps a failing of the poem, but what is also interesting is that only 13 of the 50 participants (26%) in the prose group mentioned modelling, even though this is stated several times in the original abstract. Similarly, as can be seen from Table 2, only two of the respondents from the poetry group reference the ‘Future’, whereas 14 of the participants from the prose group make reference to this fact. Whilst it could be argued that ‘Climate change’ might imply a future event, the respondents in the ‘Future’ category made explicit reference to a future scenario. Given that the poem talks about what may happen in the future, it was surprising to see that only 4% of the respondents picked up on this. Similarly, the fact that only 28% of participants from the prose group made reference to this was lower than might have been expected.\n\nRegarding the ‘Location’, ‘Impacts’, ‘Ice’ and ‘Environment’ categories, the proportion of respondents was almost identical. Of these, the relatively low number of responses in the ‘Location’ category from both the poetry and the prose groups (20% and 18%, respectively) was perhaps the most surprising, as both abstracts are very explicit in their description of Canada being the location of this study. It is perhaps not surprising that there are more references to ‘Climate change’ in the poetry group than in the prose group (38% compared to 26%), as the poem uses this phrase explicitly, whereas it is only implied in the original abstract. What is more unexpected is that the poetry group make more reference to ‘Global warming’ than the prose group (30% compared to 10%), even though the phrase itself appears in neither version of the abstract. Perhaps it is certain words in the poem like ‘barren’ and ‘melt’ that illicit this response. It is also surprising that there is such a large difference between the two groups in terms of the ‘Glaciers’ category, with 78% of the prose group including mention of them in their response to Q4, compared to only 36% in the poetry group, given that this term is used in both versions. However, this is probably explained by the fact that the prose version of the abstract mentions the word ‘glacier’ or ‘glaciers’ seven times, in comparison to the solitary use of the word ‘glaciers’ in the poem.\n\nIn addition to the categories that are shown in Table 2, there are also some individual responses that are worth noting. Out of all of the responses, only one respondent, from the poetry group, had no comment. Similarly, only one respondent, this time from the prose group, commented that it was unclear from the abstract what the research was about. This would seem to indicate that despite the participants not necessarily being experts in this field, their scientific background meant that they were suitable subjects for the study. Two respondents from the poetry group responded at a meta level, with one simply writing ‘poetry’ in their response to Q4, and the other making reference to the grammar of the poem. In regards to emotive responses, only one of the respondents from the poetry group made reference to this, noting that the research was about “Melting Canadian glaciers, with projections for the future (if slightly emotive!)”\n\nAs well as not conveying some key elements of the study (as was the case with the ‘Model’ category) there is also the danger that the poem might illicit in the reader an implied meaning, which is present in neither the prose abstract nor the underlying research itself. However, it is encouraging that there were no major categories that were exclusive to the poetry group alone. The results of this qualitative analysis would therefore seem to suggest that on this occasion the poem did a good job in conveying a similar meaning to that presented in the original abstract, other than the omission of the modelling aspect of the research. It is worth noting that this category was also absent from the vast majority (74%) of the responses from the prose group.\n\nThe quantitative and qualitative analysis would seem to suggest that scientists find a well-written abstract written in prose format to be more accessible than its poetic equivalent. Similarly, they are more likely to write longer, and more detailed summaries about what they understand the research to be about. However, these summaries were found to be fairly similar between the two groups.\n\nOne final comparison that is worth noting is the length of the abstracts themselves. The poem is 58 words long, whereas the original abstract consists of 204 words. Could it be that the longer length of the prose abstract, combined with the expectancy of the readers in terms of what a scientific abstract should look like, meant that the prose format was able to convey more information and was therefore more accessible, encouraging more verbose and detailed summaries from the participants?\n\n\nConclusion\n\nThis study presented itself as an investigation into whether poetry could be used as an alternative form of abstract in scientific journals. The rationale was that poetry might be a more effective and easily accessible format in terms of communicating the scientific findings to a general audience, but that in order for this methodology to be considered it was first necessary to try and determine if replacing a traditional abstract with a poem would still be useful to scientists who were reading the article.\n\nThe quantitative analysis of the survey suggests that whilst a scientific audience find a poetic interpretation of a scientific abstract to be no more interesting or inspiring than the original prose, they do find it less accessible. However, from the qualitative analysis, the interpretation by the two groups was similar, with the notable exception of the importance of modelling in the scientific study, which was absent from the poem. This would seem to suggest that for future studies a more suitable approach would be for the poem to be first peer-reviewed by the author of the original scientific study, in order to make sure that there were no omissions in terms of content, or indeed any additional inferences that were present due to an overly liberal artistic license.\n\nFurther research is needed to improve the practice of communicating science (see e.g. Treise & Weigold, 2002), and by investigating alternative methods of communications it is also possible to determine which areas are effective, and which require the most improvement. For example, from the results of this study, the scientific audience found the original abstract to be reasonably accessible and interesting, but they were not particularly inspired to go and find out more about the research. Could poetry therefore be written with a focus on inducing inspiration in the reader, to be read alongside the original abstract, which would still provide the accessibility required?\n\nFrom the results of this research, a future study is now planned to investigate the effectiveness of abstracts in poetic form for a general audience. Depending on the results of that study it might be that poetic abstracts could be offered as an alternative abstract, to sit alongside the more traditional prose format. However, from a scientific point of view the results presented here suggest that poetry alone is not an effective representation of the underlying research in a scientific study.\n\n\nData availability\n\nF1000Research: Dataset 1. Answers to poetry and prose surveys, 10.5256/f1000research.7783.d111682 (Illingworth, 2016).",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe author gratefully acknowledges the participation of all of the people who took part in the study.\n\n\nSupplementary Material\n\nSurvey 1: poetry group.\n\nClick here to access the data\n\nSurvey 2: prose group.\n\nClick here to access the data\n\n\nReferences\n\nBolch T, Buchroithner MF, Peters J, et al.: Identification of glacier motion and potentially dangerous glacial lakes in the Mt. Everest region/Nepal using spaceborne imagery. Nat Hazards Earth Syst Sci. 2008; 8(6): 1329–1340. Publisher Full Text\n\nClarke GKC, Jarosch AH, Anslow FS, et al.: Projected deglaciation of western Canada in the twenty-first century. Nature Geosci. 2015; 8: 372–377. Publisher Full Text\n\nDarwin E: The Botanic Garden: A Poem, in Two Parts. Part I. Containing the Economy of Vegetation. Part II. The Loves of the Plants. With Philosophical Notes. T. & J. Swords, printers to the Faculty of Physic of Columbia College. 1798. Reference Source\n\nHalliday MAK, Martin JR: Writing science: Literacy and discursive power. Taylor & Francis. 2003. Reference Source\n\nIllingworth S: Dataset 1 in: Are Scientific Abstracts Written in Poetic Verse an Effective Representation of the Underlying Research? F1000Research. 2016. Data Source\n\nJohn C, Dana N, Sarah AG, et al.: Quantifying the consensus on anthropogenic global warming in the scientific literature. Environ Res Lett. 2013; 8(2): 24024. Publisher Full Text\n\nJohnson F: Automatic abstracting research. Libr Rev. 1995; 44(8): 28–36. Publisher Full Text\n\nKahan DM, Jenkins-Smith H, Tarantola T, et al.: Geoengineering and Climate Change Polarization: Testing a Two-Channel Model of Science Communication. Ann Am Acad Polit Ss. 2015; 658(1): 192–222. Publisher Full Text\n\nKelly R: May Day. Canada, Parsifal Press. 2007. Reference Source\n\nMarshall SJ: Meltwater run-off from Haig Glacier, Canadian Rocky Mountains, 2002–2013. Hydrol Earth Syst Sci. 2014; 18(12): 5181–5200. Publisher Full Text\n\nMeadows A: The scientific paper as an archaeological artefact. J Inform Sci. 1985; 11(1): 27–30. Publisher Full Text\n\nNisbet MC, Scheufele DA: What’s next for science communication? Promising directions and lingering distractions. Am J Bot. 2009; 96(10): 1767–1778. PubMed Abstract | Publisher Full Text\n\nOrasan C: Patterns in scientific abstracts. Proceedings of Corpus Linguistics 2001 Conference. 2001; 433–443. Reference Source\n\nPinsky R: Democracy, culture and the voice of poetry. Princeton University Press. 2009. Reference Source\n\nPiotrowski AM, Goldstein SL, Hemming SR, et al.: Temporal relationships of carbon cycling and ocean circulation at glacial boundaries. Science. 2005; 307(5717): 1933–1938. PubMed Abstract | Publisher Full Text\n\nPurtill R: The purpose of science. Philos Sci. 1970; 37(2): 301–306. Publisher Full Text\n\nRudiak-Gould P: The Influence of Science Communication on Indigenous Climate Change Perception: Theoretical and Practical Implications. Hum Ecol. 2014; 42(1): 75–86. Publisher Full Text\n\nShelley PB: A Defence of Poetry. ed. RH Shepherd, Prose Works. 1888; 2. Reference Source\n\nSwales J: Genre analysis: English in academic and research settings. Cambridge University Press. 1990. Reference Source\n\nTickell C: Communicating climate change. Science. 2002; 297(5582): 737. PubMed Abstract | Publisher Full Text\n\nTreise D, Weigold MF: Advancing Science Communication: A Survey of Science Communicators. Sci Commun. 2002; 23(3): 310–322. Publisher Full Text\n\nVuille M, Francou B, Wagnon P, et al.: Climate change and tropical Andean glaciers: Past, present and future. Earth-Sci Rev. 2008; 89(3–4): 79–96. Publisher Full Text"
}
|
[
{
"id": "12381",
"date": "26 Feb 2016",
"name": "Magnus Johnson",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI would really like to see this paper pass peer review eventually as it is undoubtedly an interesting topic. However, in my opinion, it needs a lot of work to knock it into shape and they may even change the conclusions.I think a bit more here on what an abstract should be (or is) and previous research on the clarity of abstracts would be useful. An abstract, at its most fundamental, can consist of 4 sentences:What are you interested inHow did you study itWhat did you findWhy is that interesting?It should be completely clear, written in simple language and understandable to those outside the research field and hopefully to the general public.You could take any abstract from Science or Nature and it would often not be understandable by a lay-person, or most scientists that are not active in the particular field that the paper is about. You have chosen an example that is relatively good, I wonder if you had presented folk with something a little more dense whether you would have got the same result?The dataset is much richer than suggested by your simple analyses. I would like to have seen some analysis of demographics v response - otherwise why present the data in your supplemental? Do men and women respond in the same way? What about the proximity of someone’s own topic/job/role to that of the abstract you provided. I like John Wedgewood Clarke’s poem on Marine Protected Areas 1, and find it a useful tool for contemplation but is that because it is an area close to my heart? Could the font on the boxplots be enlarged so that the figures are legible without having to download the paper/slide?The non-parametric equivalent of central tendency is a median rather than a mean. You are using the appropriate figure (boxplots) which give the median, 25 & 75% quartiles and 95% confidence interval. You then give means and standard deviations in table 1 – these are not really appropriate for non-parametric data.The probability distributions of data in different categories does impact on how you interpret your results. https://statistics.laerd.com/premium-sample/mwut/mann-whitney-test-in-spss-2.phpLots of mixing of parametric and non-parametric descriptions of data – you need to avoid that. I’m not sure there is any reason why you should assume normality of the data so non-parametric would be the logical way to go. The length of responses v prose/poetry analyses are problematic in my opinion. Does their length reflect the length of the abstract/poem? The poem necessarily contains less detail. You may have got a different response had you asked – “why is this piece of work important” where the responded had to contemplate on rather than just reflect the contents. The discussion is very weak. I would expect to see the results reflected in the literature. You may be better splitting the results and discussion section into two parts. The fact that folk responding to the poem did not use the word “Model” is hardly surprising when the word is not in or directly implied in the poem. Had the first line been “In Canada a model found” the result would undoubtedly be different.",
"responses": [
{
"c_id": "1898",
"date": "26 Apr 2016",
"name": "Samuel Illingworth",
"role": "Author Response",
"response": "I would really like to see this paper pass peer review eventually as it is undoubtedly an interesting topic. However, in my opinion, it needs a lot of work to knock it into shape and they may even change the conclusions. Thank you for your comments, which have proven to be extremely useful. Please see below for a discussion of the work that has been done in order to respond to your comments.I think a bit more here on what an abstract should be (or is) and previous research on the clarity of abstracts would be useful. An abstract, at its most fundamental, can consist of 4 sentences:What are you interested inHow did you study itWhat did you findWhy is that interesting?It should be completely clear, written in simple language and understandable to those outside the research field and hopefully to the general public. Thank you for raising this issue. I agree that more could have been written about the purpose of the scientific abstract, with the following text now included in the introduction: “Andrade (2011, pp. 172) notes that “for the vast majority of readers, the paper does not exist beyond its abstract,” with the majority of researchers using the abstract to determine if the scientific study is relevant to them and worthy of a further investment of their time in reading it in its entirety. As noted by Fletcher (1988), the creation of an abstract is often also an extremely important process for clarifying the narrative of the scientific study in the mind of the author(s) themselves. Hartley (2003) also found that structured abstracts (i.e. those split into subheadings of: Background, Aims, Methods, Results, and Conclusions, or their equivalents) were found to be more informative and also provide greater clarity than their unstructured counterparts. Whilst the exact format and structure of the abstract will be determined by the journal in which the article is to be published, the purpose of writing an abstract should be to extract and summarize (Alexandrov and Hennerici, 2006), with the primary objective to not only provide information, but also to convince the reader to finish the remainder of the paper, which in some instances will involve paying for the privilege (Koopman, 1997). Orasan (2001) also observed that many authors of scientific papers do not consider the abstract to be particularly important, arguing that in many cases it is written as a necessity just before the paper is submitted. Is it therefore the case, that rather than being an effective and economical method of communicating the research, that the abstract instead represents a rushed précis of the research findings, with an even higher lexical density than that of the main body of the text? If so, then are they useful to anyone who might consider themselves, or indeed be considered an outsider? And if non-experts are unable to fully grasp the summary of the explanation, then what hope do they have of being able to fully understand the research and its potential relevance to them? Cross and Oppenheim (2006) also notes that there is probably little formal training in abstract writing, which is why in some instances there may be a lack of clarity in the abstracts that are produced in scientific journal articles.” I have specifically tried to not advise on the layout of an abstract, because as mentioned in the above text this is something that is specific to the journal. However, I hope that the above addition, as well as the additional information in regards to the primary objective of the scientific abstract seek to better highlight previous work that has been done in relation to the role of the abstract.You could take any abstract from Science or Nature and it would often not be understandable by a lay-person, or most scientists that are not active in the particular field that the paper is about. You have chosen an example that is relatively good, I wonder if you had presented folk with something a little more dense whether you would have got the same result? I agree that the result may have been different if a different article had been presented to the participants. However, this is beyond the scope of this study, and in this instance the abstract was specifically chosen so as not to negatively bias the study against the original prose. This is explained further in the revised manuscript, with the inclusion of the following text in the Survey Selection section: “It was also important that the abstract that was chosen was itself well written, and that it met the primary objectives of an abstract that was described in the Introduction, i.e. that it presented a clear and accurate summary of the paper, and left the reader wanted to find out more. Whilst this latter point is relatively subjective, it was the author’s opinion that this abstract did indeed meet these primary objectives, thus the reason for its selection. If a less clear or less obviously enticing abstract had been chosen then there was a risk that the study might be being perceived as negatively biased towards the prose version of the abstract. It is acknowledged that in choosing such an effective abstract the study might instead by positively biased towards the original prose, but given the nature of the investigation it was felt that this was more appropriate.”The dataset is much richer than suggested by your simple analyses. I would like to have seen some analysis of demographics v response - otherwise why present the data in your supplemental? Do men and women respond in the same way? What about the proximity of someone’s own topic/job/role to that of the abstract you provided. I like John Wedgewood Clarke’s poem on Marine Protected Areas 1, and find it a useful tool for contemplation but is that because it is an area close to my heart? Thank you for this comment. The data was presented in the supplementary analysis because I wanted it to be made available to encourage other researchers to use it, as I agree that there is some demographic data there that may be useful for a future study. However, given the limitations of this particular study (which are now more explicitly discussed in the Conclusions), I don’t believe that it would add anything further to the analysis of the data, in terms of the expectations that were set out in the Introduction. As explained in the response to Referee 1, this study was carried out without any funding, and part of its purpose was to demonstrate that there is an interest and a capacity for such investigations, thereby hopefully generating both future funding and also inspiration for further studies.Could the font on the boxplots be enlarged so that the figures are legible without having to download the paper/slide? Thank you for pointing this out, I have now enlarged the font considerably.The non-parametric equivalent of central tendency is a median rather than a mean. You are using the appropriate figure (boxplots) which give the median, 25 & 75% quartiles and 95% confidence interval. You then give means and standard deviations in table 1 – these are not really appropriate for non-parametric data.The probability distributions of data in different categories does impact on how you interpret your results.https://statistics.laerd.com/premium-sample/mwut/mann-whitney-test-in-spss-2.php Lots of mixing of parametric and non-parametric descriptions of data – you need to avoid that. I’m not sure there is any reason why you should assume normality of the data so non-parametric would be the logical way to go. Thank you for pointing this out, and I apologize for any confusion that was brought about by this mixing. All of the statistics in Table 1 have been replaced with median values, as recommended. In addition to that the text now reflects the non-parametric analysis that was carried out. I have left in the information relating to the Shapiro-Wilk test, as I believe that is important to statistically demonstrate that a non-parametric approach was appropriate, so that all readers could appreciate this. The length of responses v prose/poetry analyses are problematic in my opinion. Does their length reflect the length of the abstract/poem? The poem necessarily contains less detail. You may have got a different response had you asked – “why is this piece of work important” where the responded had to contemplate on rather than just reflect the contents. Thank you for raising this point, as it is important that it is addressed in the text. I believe that the brevity in the responses to the poem (in comparison to the prose) may in part be due to the participant’s relative lack of expertise in analyzing poetry. Just because a poem is shorter does not necessarily mean that it contains less detail, although this detail might not necessarily be explicit. I have now raised this issue in the text with the following commentary: “Could it be that the longer length of the prose abstract, combined with the expectancy of the readers in terms of what a scientific abstract should look like, meant that the prose format was able to convey more information and was therefore more accessible, encouraging more verbose and detailed summaries from the participants? Could it also be that the participants were on the whole less experienced in analysing poetry, and therefore felt less confident in elucidating on their opinions as to the nature of the poem? Whilst the poem is shorter than the prose and contains less statistical information, does that necessarily mean that it contains less detail? Could it be that instead of explicitly communicating detail (as is the case with the prose abstract), the poem instead had the affect of implying detail via an emotive response or reflection by the reader? That only one of the participants commented on the emotive nature of the poem suggests that in this instance it may not be the case, and that for the participants of this study there probably was a perceived lack of detail in the poem compared to the prose. However, as this was not commented on (nor asked by the survey) explicitly this cannot be confirmed.”The discussion is very weak. I would expect to see the results reflected in the literature. I respectfully disagree with this comment. I think that a detailed discussion has been given which both analyses the quantitative and qualitative data, and addresses the aims that were set out in the Introduction. I was unable to find any literature that would have been appropriate to compare to in the discussion, as I believe that this study is the first of its kind. You may be better splitting the results and discussion section into two parts. I think that in this instance it is better to group the results and discussion into one section, as the current layout of the paper allows a stronger narrative to be told, with the quantitative analysis leading onto a presentation of the qualitative data and then an analysis of this. The fact that folk responding to the poem did not use the word “Model” is hardly surprising when the word is not in or directly implied in the poem. Had the first line been “In Canada a model found” the result would undoubtedly be different. Thank you for raising this issue, as it was not properly addressed in the original manuscript. The following text has now been added to the end of the discussions section, which also includes a contextualization of the issue of transformation that is now addressed following a recommendation by Referee 1: “As discussed in the Introduction, the very nature of this study involves transforming the abstract in some way, and whilst every effort has been made in this transformation to retain the information of the original abstract, it is clear from an analysis of the surveyed responses that this has not been entirely successful. This is most clearly evidenced with the omission of the word ‘model’ from the poem. Whilst it is likely that the results would have been different had the first line of the poem been replaced with “In Canada a model found,” this is not the transformation that the author decided upon. Has the text’s primary goal therefore also changed? As discussed in the Introduction, the primary goal of the scientific abstract is to offer an effective summary of the study, but also to compel the reader, in this instance to read the rest of the article. The results of the survey would suggest that rather than a transformation of primary goal, there has instead been a transformation of focus. As with the original abstract, the purpose of the poem was still to inform and entice scientists, by presenting an indicative summary of the underlying research, and also serving as a compelling case that the remainder of the article was worthy of the reader’s attention. The analysis of the qualitative data would seem to suggest that the poem has still served that purpose (seemingly neither improving nor diminishing the desire to find out more about the study), however the focus of the summary has undoubtedly shifted.”"
}
]
},
{
"id": "12267",
"date": "08 Mar 2016",
"name": "Katherine Rowan",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI enjoyed reading this paper because it is clearly written and its core question, how should we make science accessible and compelling, matters. I also appreciate knowing about the author’s blog. The author’s argument is that it is possible to transform a scientific abstract into accessible and compelling poetry and that the poetic version can be effective, accessible, and compelling. I think he offers a demonstration that suggests this may be possible. However, this argument is somewhat like saying you can transform a tractor into a bicycle and still retain many of the tractor’s key features. You can, but the process of re-engineering the tractor as a bicycle changes not only its form but also its function. Similarly, re-writing an abstract as poetry is not a translation. It’s a transformation. The author transformed the goal of the abstract when he rendered it as poetry. I think this paper should pass peer review, if it discusses the goals that animate science versus the goals that animate popularization. Here are some resources that may be useful for this discussion. Jane Martin (1970) distinguishes explanation to prove (science, scholarship) from explanation to teach (popular science, textbooks). James Kinneavy (1971) makes a similar argument, saying scholarship and journalism are both forms of reference discourse, text designed to represent reality. Scholarship is primarily designed to present proof or evidence for its claims; journalism and popularizations aim to make research accessible and compelling. In Kinneavy’s theory, poetry is viewed as literary discourse, text designed to edify and entertain. I extend Kinneavy. In a series of papers, I (Rowan, 1988, 1989, 1990, 2003) distinguish texts designed to create awareness (headlines, weather reports, sport scores) from texts that deepen understanding (explanatory science features, textbooks).\n\nWhen Illingsworth casts a scientific abstract as poetry, he changes the text’s primary goal. It no longer is focused on offering careful evidence for a claim, but on making the text’s claim accessible and compelling. There should be a section in this paper noting this shift. One other limitation to the study is its design. Currently, the design involves a sample of one: one study. Ideally, there should be 75 or so studies and poetic presentations of their abstracts. This ideal design would avoid confounding the effects of this one study with the effects of interest (does a poetic account make a scientific abstract more accessible and compelling). Realistically, Illingsworth has already done substantial work in presenting the evidence he did concerning one abstract and its poetic version’s effects. Perhaps three to five abstract-poetry pairs would be a more realistic request for the study’s design. Then, there would be a chance to observe whether the topic of a study has an impact on the chances of rendering its abstract in poetry. Regardless, I think this paper is so well written and its topic so important that I hope a revised version, with the limitations or concerns noted here, is published.I enjoyed learning about Illingsworth’s work and will share his paper with my students in science journalism.",
"responses": [
{
"c_id": "1897",
"date": "26 Apr 2016",
"name": "Samuel Illingworth",
"role": "Author Response",
"response": "I enjoyed reading this paper because it is clearly written and its core question, how should we make science accessible and compelling, matters. I also appreciate knowing about the author’s blog. Thank you very much for these kind comments. I am very glad that you enjoyed reading the paper, and that you found the subject matter to be important.The author’s argument is that it is possible to transform a scientific abstract into accessible and compelling poetry and that the poetic version can be effective, accessible, and compelling. I think he offers a demonstration that suggests this may be possible. However, this argument is somewhat like saying you can transform a tractor into a bicycle and still retain many of the tractor’s key features. You can, but the process of re-engineering the tractor as a bicycle changes not only its form but also its function. Similarly, re-writing an abstract as poetry is not a translation. It’s a transformation. The author transformed the goal of the abstract when he rendered it as poetry. This is an excellent point, and the issue of transformation is discussed in more detail in the response to the following paragraph.I think this paper should pass peer review, if it discusses the goals that animate science versus the goals that animate popularization. Here are some resources that may be useful for this discussion. Jane Martin (1970) distinguishes explanation to prove (science, scholarship) from explanation to teach (popular science, textbooks). James Kinneavy (1971) makes a similar argument, saying scholarship and journalism are both forms of reference discourse, text designed to represent reality. Scholarship is primarily designed to present proof or evidence for its claims; journalism and popularizations aim to make research accessible and compelling. In Kinneavy’s theory, poetry is viewed as literary discourse, text designed to edify and entertain. I extend Kinneavy. In a series of papers, I (Rowan, 1988, 1989, 1990, 2003) distinguish texts designed to create awareness (headlines, weather reports, sport scores) from texts that deepen understanding (explanatory science features, textbooks). When Illingworth casts a scientific abstract as poetry, he changes the text’s primary goal. It no longer is focused on offering careful evidence for a claim, but on making the text’s claim accessible and compelling. There should be a section in this paper noting this shift. I think that this is an excellent point, thank you so much for bringing this to my attention, and also for providing some very useful and extremely interesting references. The issue of the goals which animate science vs. those which animate popularization need to be discussed, and are presented in the introduction as such: “It is also important to consider the issue of transformation, and how this potentially affects the goal of the abstract. As discussed above, the primary objective of the abstract is to both provide information and also to convince the reader to finish the remainder of the paper. The nature of the poem means that it is more likely to be thought of as a popular piece of science writing rather than a professional piece of science writing, as is the case for the original abstract. As such, care must be taken to ensure that in the transformation from prose into poetry, the primary objective does not also fully transform into that of purely establishing the novelty of the topic (Rowan, 1989). In order for the poem to still be useful to scientists it must still provide a useful summary of the research. However, could the poem do so in a more accessible manner than that of the more traditional scientific abstract?” However, it is also worth noting that the primary goal of the scientific abstract is also to some extent to entice the reader, and to encourage them to read the rest of the article. Therefore, whilst the primary objective of the scientific article might be offering careful evidence for the claim, the primary objective of the scientific abstract is to offer an accurate summary which also compels the reader to continue to the rest of the article. The primary objectives of the abstract are discussed in more detail in the revised Introduction to the manuscript, as discussed in the response to Referee 2. Whilst I agree that I have transformed the abstract by presenting it as a poem, I did so in such a way that I tried to still preserve the scientific sense of the article in question. This study was also focused on whether or not researchers would still find the poem to be accessible in terms of how useful it was in conveying scientific information to them, and as I outlined in the Introduction, the purpose of this study was to see if a poem was still “accessible and informative to other scientists.” As such I believe that there was a shift, but maybe not to the extent that you have outlined above. However, the fact that there is a shift at all is a very important point, and as such has now been addressed in the discussion, with the addition of the following text: “As discussed in the Introduction, the very nature of this study involves transforming the abstract in some way, and whilst every effort has been made in this transformation to retain the information of the original abstract, it is clear from an analysis of the surveyed responses that this has not been entirely successful. This is most clearly evidenced with the omission of the word ‘model’ from the poem. Whilst it is likely that the results would have been different had the first line of the poem been replaced with “In Canada a model found,” this is not the transformation that the author decided upon. Has the text’s primary goal therefore also changed? As discussed in the Introduction, the primary goal of the scientific abstract is to offer an effective summary of the study, but also to compel the reader, in this instance to read the rest of the article. The results of the survey would suggest that rather than a transformation of primary goal, there has instead been a transformation of focus. As with the original abstract, the purpose of the poem was still to inform and entice scientists, by presenting an indicative summary of the underlying research, and also serving as a compelling case that the remainder of the article was worthy of the reader’s attention. The analysis of the qualitative data would seem to suggest that the poem has still served that purpose (seemingly neither improving nor diminishing the desire to find out more about the study), however the focus of the summary has undoubtedly shifted.” The issue of transformation was also further addressed in the conclusions, with the following text: “The issue of transformation that was raised in the discussions is notable, and is certainly worthy of further investigation in future studies. However, such studies would need to be designed so as to specifically address this issue. For example, participants could be asked for their interpretations of the primary objective of each version of the abstract (poetry or prose). Alternatively, participants could be shown both versions of the abstract and could be asked to comment on similarities and differences between the two, both literally and in terms of what they convey.” One other limitation to the study is its design. Currently, the design involves a sample of one: one study. Ideally, there should be 75 or so studies and poetic presentations of their abstracts. This ideal design would avoid confounding the effects of this one study with the effects of interest (does a poetic account make a scientific abstract more accessible and compelling). Realistically, Illingworth has already done substantial work in presenting the evidence he did concerning one abstract and its poetic version’s effects. Perhaps three to five abstract-poetry pairs would be a more realistic request for the study’s design. Then, there would be a chance to observe whether the topic of a study has an impact on the chances of rendering its abstract in poetry. I agree that the sample size is small. However, throughout the paper I have ensured that I have not made any generalizations, and I have acknowledged that there are clearly limitations in what can be derived from the study. Whilst conducting 75 studies would certainly yield more conclusive results, it is beyond the scope and indeed the resources of this study, which was carried out without any funding. Part of the purpose of this study was to demonstrate that there is an interest and a capacity for such investigations, which can hopefully be used as leverage for further research, in which for example the issue of transformation can be more specifically addressed. Regarding the use of three to five different poems and abstract comparisons, again I think that this would make for an interesting study, but that it would be beyond the scope and resources of this investigation. Furthermore, and pending future funding, I think it would potentially be more useful to repeat the study with three to five different groups of scientists using the same poem, as otherwise it may be difficult to analyse the data across the different poems. For example, all poems would probably have to be written in a fairly similar style and using a similar metric and structure, as otherwise there may be different reasons for the way in which the poem was ‘received’ by the participants. However, this is something that could also potentially be investigated in future studies, as could a comparison of different poems written by different authors, all of whom might have been given a different remit to perform their transformation under. There is a genuine wealth of possibilities that are available here for investigation, and I hope that this paper acts as a starting point for such studies, both by myself and ideally from other interested colleagues as well. Regarding the limitations of the study I have made further explicit reference to them with the following passage in the conclusions: “It is acknowledged that the results and subsequent analysis that are presented here represent the responses from only one study that was carried out on a small subset of participants. As such, it is important to recognise the limitations of the findings, and to allow that a different set of results may be evident if a different group of participants were surveyed. Likewise, the responses of the participants would probably be different if they were shown different poetical interpretations (from the same or different authors) of the same abstract. Given these limitations, I hope that this study has demonstrated that there is a capacity and a merit in such investigations, and also that it has served as inspiration for future work. ” Regardless, I think this paper is so well written and its topic so important that I hope a revised version, with the limitations or concerns noted here, is published. Thank you. I hope that I have now addressed these issues in an acceptable manner.I enjoyed learning about Illingworth’s work and will share his paper with my students in science journalism. Thank you, that is greatly appreciated."
}
]
}
] | 1
|
https://f1000research.com/articles/5-91
|
https://f1000research.com/articles/5-2076/v1
|
25 Aug 16
|
{
"type": "Research Article",
"title": "Invariant death",
"authors": [
"Steven A. Frank"
],
"abstract": "In nematodes, environmental or physiological perturbations alter death’s scaling of time. In human cancer, genetic perturbations alter death’s curvature of time. Those changes in scale and curvature follow the constraining contours of death’s invariant geometry. I show that the constraints arise from a fundamental extension to the theories of randomness, invariance and scale. A generalized Gompertz law follows. The constraints imposed by the invariant Gompertz geometry explain the tendency of perturbations to stretch or bend death’s scaling of time. Variability in death rate arises from a combination of constraining universal laws and particular biological processes.",
"keywords": [
"Mortality",
"nematodes",
"cancer",
"Gompertz distribution",
"probability theory"
],
"content": "Introduction\n\nThe coil of a snail’s shell expresses the duality of constraint and process. The logarithmic spiral of growth constrains overall form. Particular snails modulate the process of shell deposition, varying the parameters of the logarithmic spiral. To interpret the variety of snail shells, one must recognize the interplay between broad geometric constraint and the special modulating processes of individual types1.\n\nThe pattern of death in populations follows the same duality of invariant geometric constraint and modulating process. The invariant geometry of death’s curve arises from the intrinsic order of large samples2,3. A large sample erases underlying randomness, preserving only invariant aggregate values4.\n\nI extend the large-sample concept to clarify the invariant geometry of death. I then illustrate the role of particular biological processes in modulating death’s curve: the stretch of death’s time in nematode response to physiological perturbation5 and the curvature of cancer’s time in response to genetic perturbation6,7. The consequences of particular biological perturbations can only be understood within the geometry that constrains change to follow invariant contours.\n\nTo restate the puzzle: How can we relate small-scale molecular and physiological process to population consequence? The problem remains unsolved. Finch and Crimmins8 emphasized: “A key question is how to connect … [linear] aging processes to the exponential rates of accelerating mortality that set life spans. … Although we can readily assess molecular aging, such biomarkers of aging are rarely robust as predictors of individual morbidity and mortality risk in populations.”\n\n\nRandomness and invariance\n\nI begin with the relation between small-scale randomness and large-scale order. The classical theory derives from the principles of statistical mechanics2, later developed through aspects of entropy and information4,9. Here, I briefly summarize my own extension of classical results based on geometric principles of invariant measurement and scale10–13. I then show how the abstract geometry constrains the relation between biological process and the pattern of death’s curve.\n\nTo understand the probability of dying at a particular age, we begin with the geometry of probability patterns13. For an underlying quantity, z, the probability of observing a value near to z is the rectangular area with height qz, width d𝜓z, and area qd𝜓. A probability pattern is a curve with coordinates (𝜓z, qz) defined parametrically with respect to z. For the curve of death, the input, z, may be age or time.\n\nTwo invariances constrain the geometry of probability curves. First, total probability is invariantly one. Invariant total probability implies that the height of the probability curve has a natural exponential expression13\n\nIn general, we seek metrics for which it does not matter where we set our zero reference point. In geometry, a circle shifted in space retains its invariant form. Similarly, proper geometric scaling for probability patterns is shift invariant. In terms of death, any transformation of time, z, into a fundamental time metric for probability pattern, Tz, must measure time such that a shift a + Tz does not alter death’s curve. That shift-invariant requirement leads to the exponential expression13 in equation 1.\n\nThe second key invariance is that a uniform stretching or shrinking of the fundamental metric does not alter probability pattern13\n\nTo summarize, probability curves remain invariant to shift and stretch of the fundamental metric, Tz, such that\n\nAffine invariance leads to probability pattern described by a sequence of rectangular areas\n\n\nConsequences of affine invariance\n\nHere, by emphasizing the fundamental invariances, we can take the next key step in understanding the geometry of probability patterns and the curves of death. In particular, each successive application of the affine transformation (equation 3) to T leaves the probability pattern unchanged, defining an invariant group of metrics11\n\nTo find the proper metric, T, for a particular probability pattern, we only need to find the proper base scale w for which the probability pattern is shift invariant. If, for example, z is time or age, then we only need to discover the scaling, w(z), for which\n\nEquation 5 expresses the abstract form of common probability patterns11. The abstraction does not specify the two key scaling relations 𝜓(z) and w(z) that define the coordinates of the parametric probability curve (𝜓, q) with respect to z. However, the invariances that define the geometry do impose strong constraints, leading to a limited set of forms for almost all of the commonly observed probability patterns11–13.\n\nWe have two scaling relations 𝜓 and w, but only a single parametric probability curve (𝜓, q) with associated probability q d𝜓 in each increment. Thus, many different scales can express the same probability pattern. For each application, there is often a natural scale that has a simple, understandable form for its scaling relations.\n\n\nThe invariant ticking of death’s clock\n\nA natural scale corresponds to an additional invariance with a simple interpretation. That additional invariance sets the underlying metric for the pair of scaling relations. For death, we can set the probability of dying to be invariant in each increment of the scale, d𝜓, so that 𝜓 represents the uniform metric of mortality—the invariant ticking of death’s clock. This uniform metric extends the theory of extreme values and time to failure14–17 to a more abstract and general understanding of the invariances that shape all of the common probability patterns11–13.\n\nInvariant probability in each increment can be written as q d𝜓 = −dq and thus 𝜓 = − log q, in which dq is a constant incremental fraction of the total probability. I use a minus sign as a convention to express the total probability as declining with an increase in 𝜓.\n\nWith regard to dying, we may think of the total probability of being alive as declining by a constant increment of death, −dq, in each increment d𝜓. In classic epidemiology, this definition of q would be expressed as q(z) ≡ S(z), in which S(z) is the probability of survival to time z. However, it is important to consider the classic definition as a special case of the deeper abstract geometry, which leads to a more general understanding of the constraints that shape death’s curve.\n\n\nUniversal Gompertz geometry\n\nGiven the exponential form for qz in equation 2, a constant probability q d𝜓 in each increment requires d𝜓 = dT. Using the general form of T in equation 4, we have dT = eβwdw = T′dw, in which T′ > 0 is the derivative of T with respect to w. With q̂dw = q dT for the constant probability in each increment, we have\n\nThis probability pattern is expressed on the scale w, in which w defines the natural shift-invariant metric. In other words, for some underlying observable value z, such as time or age, w(z) transforms z to a scale, w, that expresses an invariant total probability q̂ dw in each increment, and for which shifts in the scale w ↦ α + w, do not change the probability pattern.\n\nThe probability pattern in equation 6 has the familiar Gompertz form. I derived that form solely from a few simple geometric invariances. The simple invariances elevate the generalized Gompertzian form to a universal geometric principle for probability patterns11,13. By contrast, the Gompertzian pattern is usually derived from descriptive statistics or from particular assumptions about processes of failure or growth.\n\n\nPattern on the observed scale\n\nWe may express the probability pattern on the scale of the underlying observable value, z. For that scale, dw = w′dz, in which w′ > 0 is the derivative of w with respect z. The abstract Gompertzian geometry in w becomes the explicit form with respect to the directly measured value z as\n\n\nThe hazard of death\n\nOnly living individuals can die. Thus, the hazard of death is the probability of dying in an incremental metric of time divided by the probability of being alive. The incremental metric scale, 𝜓z, transforms the observed value, z, which may be time or age, into the abstract incremental scale, d𝜓. The abstract expression for the hazard of dying in an increment d𝜓 is\n\nIn each increment, the probability of dying is q d𝜓. The integral in the bottom is the sum of the probabilities of dying over the period from a starting point until the current period, in which the time metric is described by 𝜓(z).\n\nThree different metrics transform the observable time input, or other measurable input, z, into the scale of analysis: T, w, and z itself. Those three metrics yield three equivalent forms for the hazard, each emphasizing different aspects of the underlying geometric invariances\n\nWe know the scale of death’s curve when we can transform our underlying observation, z, such as age, to the affine-invariant scale, T. Often, adding a constant to age or multiplying age by a constant, z ↦ a + bz, changes the pattern of death’s curve, so using age itself as the metric is usually not sufficient. We must find some transformation of age.\n\nThe middle expression in equation 10 describes the generalized Gompertzian geometry in the most direct way. In this case, when we transform z ↦ w, changing an observation such as age, z, to the metric, w(z), we only require that death’s curve be invariant to a shift, w ↦ a + w.\n\n\nThe force of death and the curvature of time\n\nWe are partitioning the scaling of death’s curve into two steps, z ↦ w ↦ T. Once we have the shift-invariant scaling of time, w, then T = eβw changes w into the ultimate affine-invariant scaling, T. To make that last change, we need to know β, which is\n\nThe expression T″ is the acceleration, or absolute curvature, of T. The expression T′ is the rate or velocity at which T is changing. Thus, T″/T′ can be thought of as the acceleration relative to the velocity.\n\nAcceleration, curvature and force are ultimately equivalent. In terms of death, for a given velocity or rate at a particular age, T′, the value of β is the relative force that bends death’s curve. The bending of death’s curve may also be described as\n\nThe invariant geometry of death’s curve in equation 12 may be expressed as a balance, β − 𝒜 = 0, between force and acceleration. That balance is roughly analogous to Newton’s second law of motion, F = m𝒜, relating force to acceleration.\n\n\nInference\n\nThe invariant geometry does not tell us the form for the shift-invariant scaling of death’s time, w, or the value of the invariant force, β, that bends death’s curve. However, the invariances strongly constrain the likely form of death’s curve and the meaningful metrics of death’s time. Importantly, these expressions allow us to transform data about rates or motions into expressions that emphasize force and causal interpretations18,19. In biology, we rarely can predict trajectories. Instead, we focus on interpreting the changes in observed trajectories with respect to hypothesized forces7,20.\n\nThe abstract geometry is correct unto itself. In application, the geometry provides a tool that we may use for particular problems. A tool is neither right nor wrong. Instead, a tool is helpful or not according to its aid in providing insight. Below I discuss some examples. A few comments prepare for the discussion.\n\nIf we knew the correct scaling for age, w(z), then within that frame of reference, the force, β, and acceleration, 𝒜, of mortality would be constant with respect to w. Thus, the frame of reference, w, provides valuable insight. However, w may turn out to be a weirdly nonlinear scaling of measured time, z, in which the form of w is difficult to determine directly. In practice, we can derive w from equation 11 by relating the hazard, h̃(z), to w by\n\nI now discuss the time scaling of mortality in nematodes and cancer. I consider these applications only to illustrate general aspects of mortality’s temporal geometry. See Stroustrup et al.5 for details about nematodes and Frank7 for details about cancer.\n\n\nNematode mortality and the stretch of time\n\nStroustrup et al.5 conclude from their study of nematode mortality:\n\n[W]e observe that interventions as diverse as changes in diet, temperature, exposure to oxidative stress, and disruption of [various] genes … all alter lifespan distributions by an apparent stretching or shrinking of time. To produce such temporal scaling, each intervention must alter to the same extent throughout adult life all physiological determinants of the risk of death.\n\nI begin with the apparent stretching or shrinking of time. I will arrive at the same description of the nematode mortality pattern as given by Stroustrup et al.5, but framed within my more general understanding of mortality’s invariant geometry. From that broader perspective, the observed stretching or shrinking of time in the nematode study can be seen as a special case of the various temporal deformations that arise with respect to mortality’s invariant scale.\n\nThe perspective of my general framing calls into question the second conclusion that each intervention must alter to the same extent throughout adult life all physiological determinants of the risk of death. I present a simple counterexample consistent with the observed patterns. My counterexample may not be the correct description of process in nematode mortality. The counterexample does, however, emphasize key aspects of the logic by which we must evaluate the relations between pattern and process in mortality.\n\nMy framework analyzes the sequence of transformations z ↦ w ↦ T. The initial input, z, typically represents what we measure, such as a standard description of time or age. We then seek a transformation, w(z), such that the parametric curve, (w, q̂), for observed or assumed probability pattern is shift-invariant with respect to w (equation 6). In other words, the shift w ↦ α + w does not alter the probability curve. When we find the shift-invariant scale for w, we have an expression for the probability pattern in terms of the Gompertzian geometry of equation 6.\n\nA probability pattern that remains the same except for a constant stretching or shrinking of time corresponds to w(z) = log z, because a constant stretch or shrink of time by a = eα yields w(az) = α + w(z). If we express the associated parametric probability curve as the relation between time and probability, (z, q̃), as in equation 7 with w = log z, we obtain a curve that is invariant to a constant stretching or shrinking of the temporal scale, z, as\n\nStroustrup et al.5 concluded that the Fréchet distribution is the best overall match to their nematode studies. However, they invoked the Gompertz-Fréchet family of distributions by appeal to traditional epidemiology and by appeal to the general form of extreme value distributions for failure times. By contrast, I derived those distributions simply as the inevitable consequence of basic assumptions about the invariant geometry of meaningful scales13.\n\n\nThe deformation of death’s time\n\nStroustrup et al.5 discussed the stretching or shrinking of death’s time by a single constant value. My framework generalizes the deformation of time in relation to death. We begin with T, the universal frame of reference for the scaling of death’s time. On the temporal scale, T, the hazard of death, h(T), remains constant at all times (equation 9). Thus, T represents the invariant ticking of mortality’s clock.\n\nGiven that universal frame of reference for time, we may then consider other temporal scales in terms of the way in which they deform the invariant frame of reference. In this case, we work inversely, by starting with T in equation 4, and then inferring the deformations with respect to the underlying scale of description, z. We can then think of the shape of the curve (Tz, z) as describing how measured time, z, is deformed in relation to the universal invariant scale of mortality’s time, Tz.\n\nIdeally, we first infer the shift-invariant scale, w(z), and then use w in equation 4 to determine the relation between T and z. In the nematode case, w(z) = log z achieved shift invariance. Thus Tz ∼ eβw = zβ. The power law curve (zβ, z), with curvature determined by β, describes the deformation of time. The different experimental treatments did not significantly alter the curvature associated with β.\n\nWe can relate increments of the measured input, dz, to increments of mortality’s universal measure, dT, by starting with equation 11 as h˜∝dT/dz, and then writing\n\nFor a case such as the nematodes in which Tz ∼ zβ, the measured temporal increments, dz, scale in relation to the universal temporal frame as dz ∝ dT/z β−1. For β > 1, measured temporal increments, dz, shrink as time passes relative to the constant ticking of mortality’s clock at dT. When we think of dT as mortality’s constant temporal frame of reference, then the deformation of measured time is\n\nThe shrinking of measured time corresponds to the increase in the rate of measured mortality, in other words, the same amount of mortality, dT, is squeezed into smaller temporal increments, dz, increasing the density of mortality per measured unit.\n\nIn other cases, the relation of measured inputs, z, to mortality’s universal scale, Tz, will have different functional forms. Those different functional forms may correspond to non-uniform stretching and shrinking of the observed temporal scale at different magnitudes of z relative to the universal frame of reference for mortality on the scale Tz. If possible, we first infer the shift-invariant scale, w(z), for example by equation 14, and then use w to determine the relation between T and z, as in the nematode example. However, in practice, it may be easier to go directly from the invariant clock, T, to the deformed time scale, z, by using the relation dz∝dT/h˜. The following critique of the conclusions by Stroustrup et al.5 about nematode mortality provides an example.\n\n\nInvariant pattern and underlying process\n\nStroustrup et al.5 claimed that all physiological determinants of the risk of death change in the same way with each experimental intervention. I present simple counterexamples. Although my counterexamples may not describe the true underlying process, they do highlight two important points. First, commonly observed patterns often express invariances that are consistent with many alternative underlying processes21,22. Second, consideration of the alternative processes with the same observable invariances leads to testable predictions about the underlying causal processes.\n\nIn these examples, suppose that death follows a multistage process, as is often discussed in cancer progression23. Following Frank7, p. 98, we may write the dynamics of progression toward mortality as a sequence of transitions\n\nAs time passes, some individuals move into later stages of progression toward death. The rate of transition from stage i to stage i + 1 is ui. The ẋ’s are the derivatives of x with respect to z. Death occurs when individuals transition into stage n. A fraction xn(z) of individuals has died at time z, and the rate of death at time z is ẋn(z) ≡ q̃, in which q̃ has the probability interpretation of equation 7.\n\nIf the transition rates are constant and equal, ui = u for all i, then we can obtain an explicit solution for the multistage model24. This solution provides a special case that helps to interpret more complex assumptions. The solution is xi(z) = e−uz (uz)i/i! for i = 0, …, n − 1, with the initial condition that x0(0) = 1 and xi(0) = 0 for i > 0. Note that the xi(z) follow the Poisson distribution for the probability of observing i events when the expected number of events is uz.\n\nIn the multistage model above, the derivative of xn(z) is given by ẋn(z) = uxn−1(z). From the solution for xn−1(z), we have\n\nAge-specific incidence is the hazard7\n\nWe can express the scaling of measured time, dz, relative to the constant ticking of mortality’s time, dT, from equation 15, by taking dT as constant and thus\n\n\nSimultaneity and temporal deformation\n\nWhen measured time, z, is small, during the initial period of the process, the deformation of time is approximately the same as the Fréchet pattern, dz ∝ 1/zn−1. This deformation in the gamma process describes the force of simultaneity. Early in the process, all components that protect against mortality remain in the initial working state. Thus, mortality requires the nearly simultaneous failure of n independent events, which creates a force that deforms the constant ticking of mortality’s clock by rescaling the measured increments, dz.\n\nAs measured time increases, the increments dz shrink, compressing the same amount of mortality, dT, into smaller measured temporal increments. As z becomes larger, the increments dz in equation 17 shrink less, because of the reduced force of simultaneity that deforms mortality’s constant clock. With larger z, the higher-power terms of the sum increasingly dominate, until the largest power term dominates and dz then ticks at a constant rate, with dz ∝ dT.\n\nThe changing deformation of dz and the associated force of mortality can be thought of roughly as follows7. Early in the gamma process, mortality requires the nearly simultaneous failure of n independent events, creating a force of simultaneity such that dz ∝ 1/zn−1. As time passes, many individuals suffer failure of some of the n processes, leaving in aggregate the equivalent of n−1 remaining protective components, and a force of simultaneity such that, approximately, dz ∝ 1/zn – 2. As more time passes, additional components fail, and the remaining force of simultaneity diminishes, until eventually only one protective component remains for those still alive, at which point dz then ticks at a constant rate, so that dz ∝ dT.\n\nWe may also express the scaling of time on the shift-invariant Gompertzian scale, w, in which β is a relative measure of the acceleration of mortality (equation 12), by using the general expression in equation 14 and the specific form of h̃ in equation 16 to yield\n\n\nAlternative models of nematode mortality\n\nWith this understanding of the gamma process, we can consider alternative interpretations of the nematode data5. I present these alternatives to illustrate the logic of mortality’s temporal scaling and the potential relation to underlying process. The data do not provide information about whether or not these alternative interpretations are the correct description of nematode mortality. The point here is that these alternatives, or some other structurally similar alternative, might be correct, and therefore the strong conclusions of the original article may be false.\n\nTo repeat the key conclusion from Stroustrup et al.5, each intervention must alter to the same extent throughout adult life all physiological determinants of the risk of death.\n\nThat conclusion is true for the simple gamma process, as summarized by equation 17. In that equation, the value of u represents the rate at which each of the n processes fails and contributes to overall mortality. If we substitute uz ↦ ξ, then the scaling of measured time, expressed as dz ∝ dξ, changes only by a constant of proportionality as the rate, u, changes.\n\nI now consider two variations on the underlying gamma process for mortality. Each of these variations leads to a constant rescaling of time, dz. However, that constant rescaling arises from underlying processes of mortality that change in different ways in response to perturbations. These examples show that the constant rescaling of time does not imply that an intervention alters to the same extent throughout adult life all physiological determinants of the risk of death.\n\nThe first example considers two distinct sets of underlying processes that influence mortality, each set composed of n processes. Mortality occurs only after the failure of all 2n processes. Before experimental perturbation, one set of n processes has a relatively slow failure rate per process of u. The other set of n processes has a relatively fast failure rate of u′ ≫ u.\n\nIn this case, the fast processes will all tend to fail early in life, almost always before all of the slow processes fail. Thus, the fast processes have little influence on mortality. The mortality rate will closely follow the gamma process with n steps, each step at rate u, as analyzed above7.\n\nNow suppose that an intervention influences all of the slow steps but none of the fast steps. The intervention changes the previously slow rate processes into fast processes, u ↦ u″ ≫ u′. After intervention, the mortality rate will closely follow the gamma process with n steps, with each step at rate u′. The mortality pattern remains unchanged, except for a constant rescaling of time.\n\nHowever, the underlying physiological processes that determine mortality have changed completely. Previously unimportant rate processes with respect to mortality now completely dominate, and previously important rate processes no longer influence mortality.\n\nThe second example considers a set of n underlying processes that influence mortality. Each process has a different failure rate of ui for i = 1, …, n, with ui < ui+1. As before, mortality occurs only after the failure of all n processes. Frank7 presented numerical studies for this heterogeneous rate process model. Typically, the faster rate processes fail early in life and have relatively little influence. The slower processes dominate the overall temporal pattern.\n\nWith n equal rate processes, the curvature declines with time, as in equation 17. With a heterogeneous set of rate processes, the curvature tends to decline more quickly as the fastest processes fail earlier, typically leaving a progressively smaller set of remaining protective mechanisms as time passes, reducing the force of simultaneity.\n\nNow suppose that the heterogeneous set of n processes has a simple hierarchy of rates, such that ui+1 = γui, in which γ > 1 is the factor by which each rate increases relative to its slower neighbor. If the only effect of an experimental intervention is abrogation of the slowest process, u1, then the hierarchy is effectively altered only by multiplying each rate in the set by γ, because the fastest processes typically have almost no influence on pattern.\n\nOnce again, the overall mortality pattern will change only by a constant rescaling of time, even though the underlying physiological processes have changed significantly with respect to their influence on mortality. In this case, the most important process that limited mortality before intervention was in effect knocked out after intervention, whereas all other processes did not change.\n\n\nDifferent processes lead to same invariances\n\nThe actual biology of nematodes will, of course, not follow exactly either of these two example cases. The examples do show, however, that a constant rescaling of measured time for mortality can arise by heterogeneous changes in the underlying physiological determinants of the risk of death.\n\nHow should we interpret the match between variants of the multistage gamma process model and the observed scaling of nematode mortality? The correct view is that the invariances expressed by the gamma model are approximately the same as the invariances that arise by the true physiological processes. Those invariances dominate the shape of the observed patterns. The examples of the gamma models are helpful, because they show the sorts of underlying processes that generate the required invariances.\n\nUltimately, the theoretical challenge is to understand the full set of underlying processes that lead to the same invariances and thus the same observed pattern22. The empirical challenge is, of course, to figure out which particular processes occur in each particular case. Success in the empirical challenge will likely depend on further progress on the theoretical challenge, because the theoretical frame strongly influences how one goes about solving the empirical problem.\n\n\nCancer incidence and the curvature of time\n\nI now turn to genetic knockouts in cancer that change the curvature of time. Cancer incidence often follows a pattern roughly consistent with a multistage gamma process7,23. Again, that match does not mean that the underlying physiological processes truly follow the assumptions of the gamma model. Instead, the correct view is that the invariances expressed by the gamma model are approximately the same as the invariances that arise by the true physiological processes.\n\nConsider the simplest gamma process, in which cancer arises only after n protective mechanisms fail. Each mechanism fails at the same rate, u. I gave the explicit solution for that process earlier. In interpreting that solution for cancer, it is important to note an essential distinction between mortality and cancer incidence.\n\nEveryone dies but only a small fraction of individuals develop a particular form of cancer. Thus, we must analyze mortality by running the measured time, or age, z, out to a large enough value so that the cumulative probability of dying approaches one. In the gamma models above, that means letting uz increase significantly above one. By contrast, if only a small fraction of the population develops cancer before dying of other causes, then we must run z only up to a time at which the cumulative probability of cancer remains small. That limit on total incidence typically means capping uz below one.\n\nWith a small maximum value of uz, the age-specific hazard simplifies approximately to h̃ ∝ zn–1, and the scaling of measured time simplifies to dz ∝ 1/zn–1. Those scalings match the Fréchet model when we equate the curvature of time, β, with the number of steps, n, and we interpret force and curvature with respect to the shift-invariant scaling of time, w(z) = log z.\n\nWe can think of n = β as the force imposed on the logarithmic time scale, log z, caused by the requirement for the nearly simultaneous failure of n protective processes. The greater n, the greater the protective force, and the greater the bending of observed time relative to cancer’s invariant clock, ticking in increments of dT.\n\nAll of that may seem to be a very abstract theory in relation to the actual physiological processes of cancer. However, certain empirical studies suggest that the simple geometric theory of cancer’s time does in fact capture key aspects of cancer’s real physiology and genetics. In particular, certain inherited genetic mutations correspond almost exactly to the predicted theoretical change in the force of simultaneity and the temporal curvature of incidence.\n\nIf a mutational knockout reduces the number of protective mechanisms by one, such that n ↦ n – 1, then the approximate pattern of incidence changes from h˜∝zn−1 to h˜∝zn−2. In other words, the force and associated curvature, β, is reduced by one.\n\nTwo classic studies of cancer incidence made exactly that comparison. Ashley25 compared colorectal cancer incidence between groups with and without an inherited mutation that predisposes to the disease. Similarly, Knudson26 compared retinoblastoma incidence between groups with and without a predisposing inherited mutation.\n\nI analyzed those same cancers with additional data that became available after the original studies6. My analysis showed that, in each case, the groups carrying the inherited predisposing mutation had a pattern of incidence that changed relative to the control groups by reducing the estimated value of β by approximately one. Thus, the genetic knockouts reduced time’s curvature by almost exactly the amount predicted by the reduced force of diminished simultaneity in the protective mechanisms.\n\n\nConclusion\n\nA few simple invariances shape the patterns of death. That geometry does not tell us exactly how biological mechanisms influence mortality. But the geometry does set the constraints within which we must analyze the relation between pattern and process.\n\nI started with the temporal frame of reference, dT, on which mortality has a constant rate, or velocity. That temporal frame, with unchanging rate, expresses the ticking of mortality’s clock in the absence of any apparent force that would change velocity.\n\nGiven that frame without apparent force, we can then evaluate other temporal scales in terms of the forces that must be applied to change mortality’s rate relative to the force-free scale. That approach focuses attention on the forces of mortality, rather than the incidence or “motion” alone, because the pattern of motion is inherently confounded with the particular temporal frame of reference18,19.\n\nMortality’s temporal frame leads to a natural expression of invariant death with respect to a universal Gompertzian geometry. That geometric expression separates the uniform application of force from the additional distortions of time with respect to observed pattern.\n\nThe examples of nematodes and cancer illustrated how to parse observable deformations of mortality’s clock with respect to invariant aspects of pattern and potential underlying explanations about process.\n\nUntil biologists can see the constraints of Gompertzian geometry on the curves of death as clearly as they can see the constraints of the logarithmic spiral on the growth curves of snail shells and goats’ horns, we will not be able to read properly the relations between the molecular causes of failure and the observable patterns of death.\n\nPut another way, geometry does not tell one how to build a bridge. But one would not want to build a bridge without understanding the constraints of geometry. Properly interpreting the duality of constraint and process with respect to pattern is among the most difficult and most important aspects of science.",
"appendix": "Author contributions\n\n\n\nSAF did all the research and wrote the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nNational Science Foundation grant DEB–1251035 supports my research.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nThompson DW: On Growth and Form. 2nd ed (Canto reprint of 1961 revised edition). Cambridge University Press, Cambridge, 1992. Publisher Full Text\n\nFeynman RP: Statistical Mechanics: A Set Of Lectures. Westview Press, New York, 2nd edition, 1998. Reference Source\n\nGnedenko BV, Kolmogorov AN: Limit Distributions for Sums of Independent Random Variables. Addison-Wesley, Reading, MA, 1968. Reference Source\n\nJaynes ET: Probability Theory: The Logic of Science. Cambridge University Press, New York, 2003. Reference Source\n\nStroustrup N, Anthony WE, Nash ZM, et al.: The temporal scaling of Caenorhabditis elegans ageing. Nature. 2016; 530(7588): 103–107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrank SA: Age-specific incidence of inherited versus sporadic cancers: a test of the multistage theory of carcinogenesis. Proc Natl Acad Sci U S A. 2005; 102(4): 1071–1075. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrank SA: Dynamics of Cancer: Incidence, Inheritance, and Evolution. Princeton University Press, Princeton, NJ, 2007. PubMed Abstract\n\nFinch CE, Crimmins EM: Constant molecular aging rates vs. the exponential acceleration of mortality. Proc Natl Acad Sci U S A. 2016; 113(5): 1121–1123. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPressé S, Ghosh K, Lee J, et al.: Principles of maximum entropy and maximum caliber in statistical physics. Rev Mod Phys. 2013; 85(3): 1115–1141. Publisher Full Text\n\nFrank SA, Smith DE: Measurement invariance, entropy, and probability. Entropy. 2010; 12(3): 289–303. Publisher Full Text\n\nFrank SA, Smith E: A simple derivation and classification of common probability distributions based on information symmetry and measurement scale. J Evol Biol. 2011; 24(3): 469–484. PubMed Abstract | Publisher Full Text\n\nFrank SA: How to read probability distributions as statements about process. Entropy. 2014; 16(11): 6059–6098. Publisher Full Text\n\nFrank SA: Common probability patterns arise from simple invariances. Entropy. 2016; 18(5): 192. Publisher Full Text\n\nEmbrechts P, Kluppelberg C, Mikosch T: Modeling Extremal Events: For Insurance and Finance. Springer Verlag, Heidelberg, 1997; 33. Publisher Full Text\n\nKotz S, Nadarajah S: Extreme Value Distributions: Theory and Applications. World Scientific, Singapore, 2000; 196. Publisher Full Text\n\nColes S: An Introduction to Statistical Modeling of Extreme Values. Springer, New York, 2001. Publisher Full Text\n\nGumbel EJ: Statistics of Extremes. Dover Publications, New York, 2004. Reference Source\n\nLanczos C: The Variational Principles of Mechanics. Dover Publications, New York, 4th edition, 1986. Reference Source\n\nFrank SA: d’Alembert’s direct and inertial forces acting on populations: The Price equation and the fundamental theorem of natural selection. Entropy. 2015; 17(10): 7087–7100. Publisher Full Text\n\nFrank SA: The inductive theory of natural selection.arXiv: 1412.1285, 2014. Reference Source\n\nFrank SA: The common patterns of nature. J Evol Biol. 2009; 22(8): 1563–1585. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrank SA: Generative models versus underlying symmetries to explain biological pattern. J Evol Biol. 2014; 27(6): 1172–1178. PubMed Abstract | Publisher Full Text\n\nArmitage P, Doll R: The age distribution of cancer and a multi-stage theory of carcinogenesis. Br J Cancer. 1954; 8(1): 1–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrank SA: A multistage theory of age-specific acceleration in human mortality. BMC Biol. 2004; 2: 16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAshley DJ: Colonic cancer arising in polyposis coli. J Med Genet. 1969; 6(4): 376–378. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnudson AG Jr: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A. 1971; 68(4): 820–823. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "15879",
"date": "14 Sep 2016",
"name": "Sean Nee",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSurvival analysis is a field with a long history, mainly associated with the name David Cox, and is of interest far outside of biology, e.g. the reliability analysis of machines. One element of it concerns making inferences about underlying mechanisms from observed processes. The historical development of the subject seems to have ended in a cul-de-sac which we see in Stroustrup's worm paper, discussed here, although clearly explored in an earlier Ricklefs paper, discussed by no-one. Frank offers an entirely new approach with a different starting point. It seems sensible to discuss the Frank paper at the junction where the new collides with the old, which is the Stroustrup worm paper, published in Nature magazine this year.\nThe heart of the matter, from my perspective, is this. The old development of the subject ended up at this point -\nSurvival data is typically well described by both Weibull and Gompertz distributions, which are described in terms of the 'hazard' function derived from them, describing your future chance of surviving beyond a particular time, given you have survived to that time so far.\n\nIf these hazards change for some reason (e.g. temperature for Stroustrup), we can model the effect of these changes in one of two ways, dreamed up by Cox because they 'work' for statistical analysis. One of these ways (AFT) is a time scaling of the hazard, in which everything happens in 'dog years' for example. The other (PH), as noted by Cox, is very hard to give a biological interpretation to in terms of underlying mechanisms, although it is very useful for prediction, evaluating drug efficacy etc.\n\nNow, the Weibull distribution can be proved analytically to allow interpretation under both these ways of changing circumstance, one of which is, itself, intrinsically hard to interpret. So if the data are Weibull, which fits the Stroustrup data, you cannot really say anything about the underlying mechanism. Unless you get very inspired and allow \"imagination\" (Ricklefs) and \"preferences\" (Stroustrup Supp Info) to assist you. This is not necessarily a bad thing, of course, but it sounds like an admission that you have gone as far as you can by the traditional route. I note that the Supp. Info Figure 1.1, which is central to the narrative flow, uses a Weibull distribution, which has a log-log hazard and the Weibull plays a large role in their statistics/simulations etc. The Gompertz has a measly log hazard, which rather screws up the simple scaling story.\nFrank now approaches the whole subject in a way which is new, thought-provoking, challenging and very welcome, starting from a rethinking of the basic meanings of statistical distributions. It is far to soon to have an opinion on the likely success of this approach in leading us out of the cul-de-sac. But I certainly have the opinion it should be published.\nThe Ricklefs paper deserves some acknowledgement and will also flag this topic for a larger audience of evolutionary ecologists, as it concerns the evolution of longevity, species variation in longevity etc. etc.\nRicklefs RE, Scheuerlein A. 2002. Biological implications of the Weibull and Gompertz models of aging. J Gerontol A Biol Sci Med Sci. 2002 Feb;57(2):B69-76. doi: 10.1093/gerona/57.2.B69",
"responses": [
{
"c_id": "2223",
"date": "10 Oct 2016",
"name": "Steven Frank",
"role": "Author Response F1000Research Advisory Board Member",
"response": "I appreciate Sean Nee's thoughtful and well written comments, which helped me to see the subject in a broader way. This report will be of interest to anyone studying patterns of mortality. I also appreciate the pointer to the Ricklefs article. On F1000Research, the referee reports are part of the final publication, so Nee's pointer to Ricklefs is now part of the published version of the article."
}
]
},
{
"id": "16246",
"date": "07 Oct 2016",
"name": "Ophélie Ronce",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper by S. A. Frank builds on his recent work on general common patterns of statistical distributions to comment on some invariance rules when examining the distribution of mortality data. In particular, the present paper shows that there always exists an adequate transformation of time (or age), such as, when measured on this scale, the shape of the survival curve is unaffected by any shift on this scale. Interestingly, Frank here also shows that on this scaling, the survival curve has necessarily a Gompertz form. On this scaling, the acceleration of mortality with (transformed) time is constant.\nThis is indeed interesting to realize that any mortality distribution would have these properties. So not much could be inferred about the mechanisms of mortality and aging just because one transformation of time would lead to the distribution of mortality satisfying those properties. The remaining question is whether we can infer something from the fact that one transformation of time rather than another produces this invariance.\nAs an illustration, the present paper by S. A Frank then comments on a recent study by Stroustrup et al. (2016, hereafter S2016), which showed that various genetic or environmental interventions affected mortality patterns in nematodes populations by modifying the time scale of mortality, leaving the shape of survival curves unchanged. More precisely, a log transformation of time then satisfies the invariance properties mentioned above. Interestingly, a few interventions in their data set do not exhibit invariance after a log-transformation of time. In the supplementary material of the S2016, the authors have examined a diversity of mechanistic models of mortality and checked under which perturbations of the parameters of these models one would observe the specific invariance pattern seen in their data. In some of these models, the specific invariance pattern emerges only when all sub-processes affecting mortality are affected to the same extent by the intervention (e.g. competing risk models), or when all parameters in the models are affected to a similar extent by the intervention (diffusion models). In some other models (e.g. network models), the specific invariance can emerge when only part of the processes are affected by the intervention. These conclusions appear therefore more subtle than the claim made in the main text of S2016 that “each intervention must alter to the same extent throughout adult life all physiological determinants of the risk of death”. It is furthermore not very clear to me what is meant exactly by “all physiological determinants of the risk of death”. Can we say that the drift and variance terms in a diffusion model for vitality correspond to different physiological determinants of the risk of death? Again we could imagine different mechanistic sub-models that would create different functional relationships between the drift and variance parameters.\nFrank uses S2016 in two ways: first, as an illustration of a particular type of invariance exhibited by real data, second, as a warning against the risk of over-interpretation of such invariances. In particular, Frank reacts to the claim cited above, which is indeed confusing. As S2016 did in their supplementary material, Frank exhibits a mechanistic model of mortality (but yet a different one from those considered S2016), which shows the same properties of invariance when only part of the processes of mortality are affected by the intervention. Both the theoretical exercise of Frank and those of S2016 show that there are many different mechanistic models that can exhibit the same specific invariance and thus raise together strong doubts about what we can infer from invariances in mortality patterns. The set of models examined by S2016 is certainly not exhaustive despite its diversity, and claims based upon this set (already more complex than the simple cited argument suggests) cannot really be generalized to all models of mortality.\nI therefore agree with the general message of caution of the present paper by S.A. Frank. It however remains frustrating that the general framework that he proposes does not really help to get a better general grasp at what features of a model would produce one type of invariance rather than another, and for instance help generating hypotheses that would explain why some interventions did fail to produce the same type of invariance in S2016. This what S2016 attempted in their supplementary material.\nIn particular, I would be interested in a clearer illustration of the claim made by Frank that “consideration of the alternative processes with the same observable invariances leads to testable predictions about the underlying causal processes”.\nInterestingly, the model put forward by Frank is shift-invariant on a scale, which is not a simple log-transformation of time (see equation 17 and that following on the right column of page 6). While early in the process, the deformation of time to achieve invariance would resemble that expected under a log-scaling, this is not true later. I therefore failed to understand exactly why this model would actually fit the data presented by S2016. More explanation would be necessary here.\nThe whole argument page 7 about interventions affecting part of the processes when slow and fast processes determine mortality makes intuitive sense but would be more convincing if illustrated (as was done for instance with examples in the supplementary material of S2016). I found these arguments about slow and fast processes to be quite disconnected actually from the general arguments about invariance presented before.\n\nTo conclude, I found the contribution of Frank novel and interesting, but still a bit frustrating about how this perspective could help us extract more information from patterns of mortality. The message of caution is an important one. The criticism of S2016 is justified by the over-simplistic claim included in their abstract and main text, but I am a bit concerned that it may misrepresent what these authors have achieved. This claim was actually motivated, not directly by the examination of invariance in the data, but by the comparison of several models to data in a spirit similar to what Frank proposes here. That several models could exhibit the same invariance rule for different reasons was clearly shown already in S2016 (supplementary material). The common properties of models exhibiting that specific invariance rule however appear even less clear than suggested by the initial exploration of S2016 and it could be good that the present paper communicates more precisely about this point.",
"responses": [
{
"c_id": "2224",
"date": "10 Oct 2016",
"name": "Steven Frank",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Ophélie Ronce has provided an excellent commentary on recent studies of mortality, placing my article in that broader context. F1000Research includes referee reports as part of the final publication, and I am very pleased to have this report included. I agree with most of the comments and opinions in this report. My primary goal was perhaps a bit simpler and less ambitious than Ronce's wider goals. I intended to show that there is a simple mathematical truth that should be kept in mind whenever one tries to understand patterns of mortality. The mathematical truth is of great value but at the same time does not tell us exactly what is happening in any particular case. I emphasized that duality very clearly in my conclusions. In the context of that duality, my discussion of Stroustrup et al. (S2016) was intended primarily as an illustration of the mathematical result rather than as an attempt to deconstruct fully their data and analyses. With regard to Ronce's specific comments: I therefore agree with the general message of caution of the present paper by S.A. Frank. It however remains frustrating ... In particular, I would be interested in a clearer illustration of the claim made by Frank that “consideration of the alternative processes with the same observable invariances leads to testable predictions about the underlying causal processes”. Perhaps my use of S2016 as an illustration was misleading. I do not view my conclusion as one of caution. Rather, there is something very simple and true about curves of death that one must keep in mind to move forward. I do not know exactly how to move forward, but I do know that I must use the truth to help focus my approach. The accomplishment is, I think, a step in the right direction, a positive rather than cautionary or negative contribution. With regard to the specific point, consider the following. Suppose that, in theory, members in one set of seemingly different processes all lead to the same general scaling of mortality. Suppose that other members of another set all lead to a different scaling of mortality, for example, a different bending of time. We now have a comparative theoretical prediction. If we can transform the underlying processes from one set to the other, the bending of time should change in a predictable way. If, by contrast, we transform from one underlying process to another within the same set, then the bending of time should not change. That is a very abstract summary, without specifying exactly what those underlying processes are. But the knowledge that such a thing is possible gives us a point of departure for moving ahead. We now know that we need better theory to tell us what we might expect, And we need to figure out how to design experiments in the context of this theory, which is much deeper and more likely to provide a way ahead than the current relatively shallow and haphazard approach. That is not a critique of S2016, which was a thoughtful article, but rather of much of the literature. But, as I said, I am not interested in caution or critique, but rather in what we need to go forward. In that regard, I agree that the supplementary material of S2016 may have provided some interesting discussion, but I think it is not what we ultimately need to move ahead. Interestingly, the model put forward by Frank is shift-invariant on a scale, which is not a simple log-transformation of time (see equation 17 and that following on the right column of page 6). While early in the process, the deformation of time to achieve invariance would resemble that expected under a log-scaling, this is not true later. I therefore failed to understand exactly why this model would actually fit the data presented by S2016. More explanation would be necessary here. That is a model of a gamma process. A gamma process is perhaps the simplest and most common model of failure. A gamma process always has that property of log scaling at small magnitudes and linear scaling at large magnitudes, leading to a transition from power law to exponential forms. I have discussed this extensively in my series of articles on probability patterns, which can be found on my web site at https://stevefrank.org. The whole argument page 7 about interventions affecting part of the processes when slow and fast processes determine mortality makes intuitive sense but would be more convincing if illustrated (as was done for instance with examples in the supplementary material of S2016). I found these arguments about slow and fast processes to be quite disconnected actually from the general arguments about invariance presented before. Such analyses would be a good project for future work. This article concerns a simple abstract mathematical point, rather than a variety of specific models tied to specific assumptions. Again, perhaps my use of S2016 was misleading with respect to my primary interest. S2016 is interesting, but only in a very limited way. On the positive side, we now know more clearly what needs to be done, so we have a basis for future work. In closing, I wish to emphasize again how much I appreciated the deep and thoughtful comments by Ronce. The overview given in this referee's report provides a better introduction to my subject than I gave in my article."
}
]
}
] | 1
|
https://f1000research.com/articles/5-2076
|
https://f1000research.com/articles/5-2061/v1
|
24 Aug 16
|
{
"type": "Research Article",
"title": "A real-world intention-to-treat analysis of a decade’s experience of treatment of hepatitis C with interferon-based therapies",
"authors": [
"Nowlan Selvapatt",
"Ashley Brown",
"Ashley Brown"
],
"abstract": "Objectives: To assess the uptake of pegylated interferon (PegIFN) plus ribavirin (RBV)-based regimens in patients with hepatitis C virus (HCV) in a large, single-centre, real-world setting over 10 years. Methods: This was a single centre, retrospective analysis of data from patients who attended their first appointment for treatment of HCV genotype 1–3 between 2003 and 2013. Patients were stratified by HCV genotype. The total number of patients who attended their first appointment, incidence of patients who did not proceed to treatment and associated reasons, and incidence of patients treated were analysed. Sustained virological response (SVR) rates were also reported for all patient populations. Results: Overall, 1,132 patients attended their first appointment; 47.8% were included in the genotype 1 group (genotype 1a: 22.2%, genotype 1b: 13.3%, genotype 1 other: 12.3%), 7.7% in the genotype 2 group and 44.5% in the genotype 3 group. A greater proportion of patients received treatment versus those who did not receive treatment (84.4% vs 15.6%, respectively). Reasons for declining treatment included: patient declined treatment with PegIFN plus RBV: 35.0%, medical contraindications: 20.3% and mental health-related contraindications: 13.6%. An SVR was achieved in 52.6% of patients who attended their first appointment and 62.3% of patients who received treatment. Conclusions: Approximately half of the patients included in this study achieved an SVR. A noteworthy proportion of patients did not receive treatment due to a reluctance to receive PegIFN plus RBV or contraindications to therapy. Results suggest an ongoing need for improvement in the treatment uptake and overall outcomes – particularly for genotype 2 and 3 patients for whom availability of interferon-free regimens is limited. The introduction of more tolerable direct-acting antiviral regimes may help overcome barriers to uptake demonstrated within this cohort.",
"keywords": [
"hepatitis C virus",
"interferon",
"ribavirin",
"real world"
],
"content": "Introduction\n\nData from the World Health Organization suggest that 130–150 million people are infected with chronic hepatitis C worldwide, a significant proportion of whom will develop liver cirrhosis or cancer1. Furthermore, the global burden of diseases, injuries, and risk factors study showed that in 2010 alone, an estimated 499,000 deaths were related to chronic hepatitis C2. The most recent estimates from the UK suggest that 214,000 people are chronically infected with hepatitis C virus (HCV) nationally; approximately 90% are genotype 1 and genotype 3 infections3.\n\nThere are seven known genotypes of HCV, although it is possible for patients to be infected with more than one genotype concurrently1,4. Treatment of HCV can be complex as the genotypes do not respond in the same way to some therapies. The armamentarium against HCV now comprises antiviral treatments that can cure approximately 90% of HCV infections, thereby reducing the risk of death from liver cancer and cirrhosis; however, global access to diagnosis and treatment remains poor 1.\n\nUntil 2011, the only approved treatment option for patients infected with HCV was a pegylated interferon (PegIFN) plus ribavirin (RBV)-based regimen administered for 48 weeks for genotype 1, and 24 weeks for genotypes 2 and 3. Sustained virological response (SVR) rates reported in the registration studies for the dual therapy, PegIFN plus RBV, were 42–52% for genotype 1 and 76–88% for genotypes 2 and 35–7. This dual therapy has been associated with frequent and sometimes serious side effects. These side effects, together with treatment durations of up to 1 year and a number of contraindications to treatment, are often seen as barriers to treatment uptake and adherence for some patients5–9.\n\nIn 2011, two first-generation protease inhibitors, telaprevir and boceprevir, were licensed for use alongside PegIFN plus RBV for patients with HCV genotype 1. This triple therapy improved SVR rates for genotype 1 patients from 42–52% to 66–75%5–7,10; however, the tolerability profiles and contraindications for use of the first-generation triple therapies remain an issue, limiting the number of patients considered suitable for treatment11. Further advances were made in the treatment options for genotype 1 patients with the introduction of IFN-free, direct-acting antiviral regimens in 2013 that have significantly improved treatment uptake, SVR rates and tolerability profiles compared with the previously available dual and triple therapies12–14. However, the availability of these IFN-free regimens is limited for treatment-naïve, genotype 2 or 3, patients in the UK.\n\nAlthough patients with HCV genotype 1 now have alternative treatment options, patients diagnosed with HCV in the real world who do not qualify for treatment with new direct-acting antivirals often decline treatment with a PegIFN plus RBV-based regimen, as they are unwilling or feel unable to endure the associated side effects. Medical and mental health-related contraindications also pose a barrier to the treatment of a proportion of the HCV-infected cohort.\n\nThis study was designed to assess the uptake of PegIFN plus RBV-based regimens in patients with chronic hepatitis C in a large, single centre, real-world setting over 10 years of treatment. SVR rates for the intention-to-treat (ITT) and treated-patient populations were compared with those achieved in randomised, controlled trials using similar treatment regimens to determine whether our real-world outcomes for patients with HCV were reflective of those achieved in randomised controlled trials.\n\n\nMethods\n\nThis study was a single centre, retrospective analysis of data from patients who were referred to, and attended their first appointment at the Liver and Antiviral Unit at St Mary’s Hospital, London (part of the Imperial College Healthcare NHS Trust), for treatment of HCV genotype 1–3 between 2003 and 2013. All treatments and follow-up appointments were also carried out in the Liver and Antiviral Unit at St Mary’s Hospital. Informed patient consent was not required as no patient identifiable information was collected and data collection was retrospective for service evaluation. The work was originally commissioned as a service evaluation by the Chief of Service for Hepatology, Professor Mark Thursz, who granted permission to use and publish the data. As part of a service evaluation, ethical approval was not required. Procedures followed were in accordance with the ethical standards of clinical treatment and within the Helsinki Declaration of 1975, as revised in 2013. It is the belief of the authors that the results of this evaluation is of interest to the wider medical community.\n\nReferred patients ≥18 years old with virologically confirmed chronic hepatitis C genotype 1–3 were eligible for inclusion in the study. Analysis of the data for patients with genotype 4 HCV have been previously published and so were not reported in this study16. The study aimed to assess all patients who were referred to the Antiviral Unit specifically for consideration of treatment by a treating hepatologist or specialist practitioner from the outpatient clinic. Therefore, all patients included in the analysis would have been seen in the outpatient setting by a specialist who intended to treat with PegIFN plus RBV. This, therefore, excluded patients who did not attend or comply with outpatient procedures or who had been deemed unsuitable for treatment by the treating physician. Patients who were referred for treatment but did not attend their first appointment at the Antiviral Unit were not included in this analysis. To ensure that the data analysed only related to patients offered an IFN-based treatment regimen (with or without first-generation protease inhibitors), patients referred for treatment after 2013 were not included in these analyses. All patients referred for treatment were screened for medical and mental health-related contraindications. Patients considered suitable for treatment were offered a PegIFN plus RBV treatment regimen over 24–48 weeks, dependent on genotype and predicted response to treatment. In 2011, when first-generation protease inhibitors became available for use in clinical practice, patients with HCV genotype 1 were offered the opportunity to include boceprevir or telaprevir in their treatment regimen.\n\nData were collected on all referrals to the Liver and Antiviral Unit using information from clinical letters and prospectively collated into a computer-based database during the study period. Database and clinical note analyses were performed to establish the total number of patients referred for treatment who attended their first appointment, the incidence of patients who did not proceed to treatment and reasons thereof, and the incidence of patients treated. The incidence of patients who achieved or failed to achieve an SVR were also reported. Analyses were undertaken on the treated patients in the genotype 1 group to establish the proportion of patients whose treatment regimen included a first-generation protease inhibitor and the SVR rates thereof. Patients were considered to have achieved an SVR if they exhibited undetectable HCV RNA 24 weeks after the completion of their antiviral therapy. All analyses were descriptive and calculations were performed using Microsoft Excel 2016 software.\n\nThe reasons given for the patients who did not receive treatment were also investigated. Patient notes were used to identify medical and mental health-related contraindications; no retrospective assessments of clinical information were carried out. Therefore, contraindications were only included if they were clearly stated in the notes by the treating medical team. When a clear reason for the patient not receiving treatment was not in the notes the reason was categorised as ‘unknown’. The cirrhotic status of the untreated patients was analysed. A patient was considered to have cirrhosis of the liver in cases where the liver biopsy ISHAK score was 5 or 6 out of 6, or the pathologist reported cirrhosis, or where a Fibroscan score was >12.4 KpA.\n\nAnalyses were undertaken using the ITT population, which included all patients who were referred for treatment and attended their first appointment at the Liver and Antiviral Unit. Analyses were repeated using the treated-patient population, which included patients who were referred for treatment, attended their first appointment and went on to receive treatment.\n\nPatients were stratified by HCV genotype; patients with HCV genotype 1a or 1b were included in their respective subgroups (genotype 1a and genotype 1b). All other genotype 1 patients, including mixed genotype and other subgroups, were included in the ‘genotype 1 other’ subgroup. The genotype 1a, genotype 1b and ‘genotype 1 other’ patient populations collectively made up the overall genotype 1 group. Patients with HCV genotype 2 were included in the genotype 2 group and patients with HCV genotype 3 were included in the genotype 3 group.\n\n\nResults\n\nA total of 1,132 patients with HCV genotypes 1–3 were referred to the Liver and Antiviral Unit for treatment between 2003 and 2013. Of these patients, 47.8% were included in the genotype 1 group (genotype 1a: 22.2%, genotype 1b: 13.3%, genotype 1 other: 12.3%), 7.7% were included in the genotype 2 group and 44.5% were included in the genotype 3 group (Figure 1, Dataset 1, Data file 1). Overall, a greater proportion of patients received treatment compared with those who did not receive treatment (84.4% vs 15.6%, respectively). A similar pattern was seen in the patient groups stratified by genotype (genotype 1: 81.3% vs 18.7%, genotype 2: 82.8% vs 17.2%, genotype 3: 87.9% vs 12.1%, respectively).\n\nOverall, a greater proportion of patients received treatment versus those who did not receive treatment.\n\nOf the 1,132 patients who were referred for treatment and attended their first appointment, 15.6% did not receive treatment. The most frequently cited reasons were patient declined treatment with PegIFN plus RBV (35.0%), medical contraindication (20.3%) and mental health-related contraindication (13.6%) (Figure 2, Dataset 1, Data file 2). These most frequently cited reasons for patients not receiving treatment remained consistent across the groups when stratified by genotype (Figure 1).\n\nThe most frequently cited reasons for not proceeding to treatment were: the patient declined treatment with PegIFN plus RBV, medical contraindication, and mental health-related contraindication. PegIFN: pegylated interferon, RBV: ribavirin.\n\nOf the patients who did not receive treatment, 17.5% had cirrhosis, 42.9% did not have cirrhosis and 39.6% did not have a cirrhotic status indicated in their notes. In the groups stratified by genotype, cirrhosis was indicated in 18.8% of the patients in the genotype 1 group, no patients in the genotype 2 group and 19.7% of the patients in the genotype 3 group. This was compared with 49.5% of patients in the genotype 1 group, 60.0% of patients in the genotype 2 group and 27.9% of patients in the genotype 3 group who did not have cirrhosis. In the genotype 1 group, 31.7% had no cirrhotic status indicated in their notes, compared with 40.0% of patients in the genotype 2 group and 52.4% of patients in the genotype 3 group (Dataset 1, Data file 3).\n\nIn this real-world study, an SVR was achieved in 52.6% of the patients who were referred for treatment and attended their first appointment. The proportion of patients who achieved an SVR was higher in the genotype 2 and genotype 3 groups compared with the genotype 1 group (63.2% and 60.5% vs 43.4%) (Figure 3, Dataset 1, Data file 4).\n\nThe proportion of patients who achieved an SVR was higher in the genotype 2 and genotype 3 groups than in the genotype 1 group. Patients were considered to have achieved an SVR if they exhibited undetectable HCV-RNA 24 weeks after completion of antiviral therapy. The ‘GT 1 other’ group included all genotype 1 patients including mixed genotypes and other subgroups that were not genotype 1a or 1b. GT: genotype, HCV-RNA: hepatitis C virus-ribonucleic acid, SVR: sustained virological response.\n\nOverall, 955 patients in this study received treatment and were included in the treated-patient population. Of these patients who received treatment 62.3% achieved an SVR. The proportion of patients achieving an SVR in the groups stratified by genotype was higher in the genotype 2 and genotype 3 groups compared with the genotype 1 group (76.4% and 68.8% vs 53.4%, respectively) (Figure 4, Dataset 1, Data file 5).\n\nThe proportion of patients who achieved an SVR was higher in the genotype 2 and genotype 3 groups than in the genotype 1 group. Patients were considered to have achieved an SVR if they exhibited undetectable HCV-RNA 24 weeks after completion of antiviral therapy. The ‘GT 1 other’ group included all genotype 1 patients including mixed genotypes and other subgroups that were not genotype 1a or 1b. GT: genotype, HCV-RNA: hepatitis C virus-ribonucleic acid, SVR: sustained virological response.\n\nProtease inhibitors were administered to 19.5% (n=86) of the treated patients in the genotype 1 group. Boceprevir was administered to 6.1% of the treated patients in the genotype 1 group and telaprevir was administered to 13.4% of the treated patients in the genotype 1 group. Overall, 72.1% of the patients who received one of these first-generation protease inhibitors achieved an SVR. Similar SVR rates were achieved with the regimens including boceprevir compared with telaprevir (74.1% vs 71.2%, respectively). Further results are presented in Figure 5 (Dataset 1, Data file 6).\n\nAn SVR was achieved in approximately three-quarters of genotype 1 patients receiving protease inhibitors. Similar SVR rates were achieved with the regimens including boceprevir or telaprevir. Patients were considered to have achieved an SVR if they exhibited undetectable HCV-RNA 24 weeks after completion of antiviral therapy. HCV-RNA: hepatitis C virus-ribonucleic acid, SVR: sustained virological response.\n\n\nDiscussion\n\nResults of this real-world, single centre, retrospective analysis of data from a 10-year period show that approximately 85% of patients who attended the Liver and Antiviral Unit for treatment of HCV received treatment. Data from Public Health England’s commissioning template for estimating disease prevalence suggest that the catchment area for the study centre (North West London boroughs of Barnet, Brent and Harrow) has an estimated 5,035 hepatitis C-infected individuals (n=1,602, n=208; n=1,504, genotypes 1–3, respectively)17. This suggests a failure to treat a large proportion of the HCV-infected population in this region when considering total treatment numbers in this centre of 955 genotype 1–3 patients from 2003–2013 and 118 genotype 4 patients from 2002–201416. Whilst this highlights issues regarding screening and patient identification, it could also reflect sub-optimal treatment uptake rates related to tolerability issues surrounding IFN-based therapies.\n\nAlthough an SVR was achieved in 62% of the treated patients, only approximately half of the patients who were referred for treatment and attended their first appointment achieved an SVR. When stratified by genotype, as expected, the proportion of patients achieving an SVR was higher in the genotype 2 and genotype 3 groups compared with the genotype 1 group.\n\nThe proportion of the treated-patient population who achieved an SVR in this study was generally in line with previously reported outcomes in clinical trials5–7. It has been suggested previously that outcomes published for treatment of HCV in clinical studies are often not reflected in real-world clinical practice9. However, this centre is a Central London teaching hospital and regional hepatology referral centre with specialist antiviral clinics and dedicated clinical nurse specialists, consultants, pharmacists and a psychiatry liaison. Therefore, it is possible that the screening and support provided at this centre enabled similar outcomes for the treated-patient population to those seen in a clinical trial-based environment. This level of resource might not be available in other centres. In this study, when the patients who attended the centre for treatment of HCV but did not receive treatment were taken into account, the SVR rates were reduced by approximately 10% across all genotypes. Furthermore, the patients who were referred for treatment and did not attend their first appointment and those with HCV who were not referred for treatment at the Liver and Antiviral Unit were not included in these analyses. The inclusion of these patients would have decreased the proportion of patients achieving an SVR further. We therefore conclude that a sub-optimal number of patients diagnosed with HCV in the UK are currently achieving appropriate treatment outcomes. These findings are in line with previously published findings by other UK-based practitioners9.\n\nMore recently the introduction of direct-acting antivirals have revolutionised chronic hepatitis C treatment with superior outcomes in genotypes 1–3 compared with IFN-based therapies18–25. The key paradigm shift, however, relates to the greater tolerability and acceptability of these treatments compared with IFN-based therapies5–7,9,18–25. In general these drugs have a narrower side effect profile, are not affected by concomitant opiate substitution and street drug use, and have fewer contraindications13,14. Taking into consideration the improved tolerability profiles and reduced medical and mental health-related contraindications with these new treatment options, we speculate that in the future a higher proportion of patients who attend the centre for treatment of chronic hepatitis C will proceed to treatment. This will, in turn, increase the overall rate of SVR when considering both a per-protocol and an intention-to-treat perspective. In addition, although a proportion of patients who receive treatment with a PegIFN plus RBV-based regimen after a relapse achieve an SVR26, it has been suggested that deferring treatment until new options are available for these patients might be preferential27.\n\nIn line with national estimates published by Public Health England, over 90% of the patients in this study had HCV genotype 1 or 3 infections3. The proportion of patients who received treatment was higher in the genotype 2 and 3 groups compared with the genotype 1 group. This could be due to known higher SVR rates and shorter treatment durations for patients with HCV genotype 2 and 3 compared with HCV genotype 16,7. Pre-2011, patients in the genotype 1 group were offered PegIFN plus RBV for 48 weeks, with a lower probability of achieving an SVR than the genotype 2 and genotype 3 groups. Towards the end of the study period, post-2011, patients in the genotype 1 group were also offered a first-generation protease inhibitor, which improved SVR rates to approximately 70%. One-fifth of the treated patients in the genotype 1 group received first-generation protease inhibitors, which raised the mean SVR rate of the patients in the genotype 1 group slightly. Approximately half of the treated patients in the genotype 1 group and three-quarters of the treated patients in the genotype 2 group achieved an SVR in this study, in line with the PegIFN plus RBV registration trials5–7,28. In this study, the proportion of patients in the genotype 3 group who achieved an SVR was slightly lower compared with the rates reported in the registration trials (69% vs 76–88%, respectively), but were in line with other European studies reporting outcomes of patients with HCV treated with a PegIFN plus RBV-based regimen5–7,9,28,29. This could be reflective of the real-world baseline demographics of this study population compared with Phase 3 clinical trial cohorts, for characteristics such as co-morbidities, age, fibrosis status, metabolic and IL-28B status.\n\nThe most commonly recorded reason for patients not receiving treatment (for over a third of patients) was a reluctance to receive a PegIFN plus RBV-based treatment regimen. This is in line with previous studies, reporting that the side effects of PegIFN plus RBV-based regimens are commonly cited as a barrier to initiation or adherence to treatment of HCV28,30. Results from a survey of treating physicians in 2010 showed that patient-related barriers, including fear of side effects, concerns regarding treatment duration and concerns regarding treatment effectiveness, were considered the most significant barrier to treatment of HCV in Western Europe31. The side effects associated with PegIFN include autoimmune syndromes, neutropenia, flu-like symptoms and neuropsychiatric disorders; while RBV has been found to induce anaemia28,30. A recent meta-analysis, including results from nine, Phase 3, clinical trials of sofosbuvir-based regimens, found that the removal of PegIFN and RBV from the treatment regimen led to a substantial improvement in patient-reported health-related quality of life during treatment32. This is in contrast with the substantial decrement in health-related quality of life and productivity reported for patients receiving a PegIFN plus RBV-based regimen32–36. Medical and mental health-related contraindications made up a further third of the reasons cited for patients not receiving treatment in this study. A US study of 45,690 HCV-infected patients reported that bipolar disorder, anaemia, pregnancy and neutropenia were the most frequently cited contraindications to PegIFN plus RBV-based therapy37. A small proportion of patients were lost to follow-up or moved location and the remaining fifth of patients who did not receive treatment had no clear reason recorded in their patient notes. An economic model from the USA analysing work productivity of patients with HCV genotype 1 compared patients treated with an all oral direct-acting antiviral (ledipasvir/sofosbuvir)-based regimen versus no treatment. Patients with untreated HCV were reported to impose a substantial societal burden due to reduced work productivity; the model predicted that the treatment of patients using a (ledipasvir/sofosbuvir)-based regimen would result in significant cost savings from a societal perspective38.\n\nApproximately a fifth of patients with HCV genotypes 1 or 3 who did not receive treatment had cirrhosis. A recent study of HCV-infected patients showed that patients with cirrhosis who achieved an SVR had a 5-year mortality rate of 5%, rising to over 15% for patients who did not achieve an SVR. After adjustment for potential confounding factors, achieving an SVR was found to be associated with an approximately 74% decreased risk of all-cause mortality in the cirrhotic cohort39. This indicates a significantly reduced risk of death for HCV-infected patients with cirrhosis who achieve an SVR, although it is important to recognise that patients with cirrhosis have been reported to show a reduced response to antiviral therapy compared to those without cirrhosis40. The recent advent of all oral direct-acting antiviral treatments has increased the capability to treat patients with cirrhosis compared to IFN-based regimens, although it is not clear yet how this may alter the natural history of disease41,42.\n\nA number of limitations should be considered for this study including the retrospective nature of the analyses and the lack of comprehensive baseline demographics. The overall treatment population included patients who attended their first appointment at the Liver and Antiviral Unit; therefore, patients who were considered eligible and referred for treatment but did not attend their first appointment, for whatever reason, were not included in these analyses. This study is likely to have significantly under-reported the number of patients overall who were not included in the intention-to-treat analysis. Given that the analyses included only those patients who were referred directly for treatment, it is likely to exclude a multitude of patients who were either not referred or self-elected not to embark on antiviral therapies before an opportunity to be referred for treatment was offered. The concern is that this group of patients may have since developed complications of the virus or remain unlinked to care and have ongoing potential to develop HCV-related complications. More tolerable and acceptable treatments, initiated at an earlier stage with less need for specialist involvement, would conceivably increase treatment uptake rates and as a result, reduce long-term disease burden. The treatment options available for patients with HCV genotype 1 changed during the study period, with the introduction of first-generation protease inhibitors in 2011; the genotype 1 data should be interpreted with this in mind.\n\nResults from this 10-year retrospective analysis of real-world data suggest that half of the patient population who attended the Liver and Antiviral Unit for IFN-based treatment of HCV went on to achieve an SVR. Furthermore, over half of the patients with HCV genotype 1 and one-third of the patients with HCV genotypes 2 and 3 failed to achieve an SVR. A noteworthy proportion of patients did not receive treatment due to a reluctance to receive a PegIFN plus RBV-based regimen or contraindications to therapy which may not be relevant to current direct-acting antiviral treatments. Whilst interferon therapies offer reasonable treatment outcomes for carefully selected patients at a population level, issues pertaining to patient perceptions and contraindications are a barrier for upscaling of treatments. Despite these major advances in the therapeutic options available for treatment of HCV, there remains an ongoing need for improvement in the treatment uptake and overall outcomes for the HCV-infected genotype 2 and 3 patients in our UK-based centre.\n\n\nData availability\n\nF1000Research: Dataset 1. Data for real world intention-to-treat analysis of a decade's experience of treatment of hepatitis C with interferon-based therapies, 10.5256/f1000research.9114.d13355943",
"appendix": "Author contributions\n\n\n\nBoth authors met ICMJE criteria for authorship. AB collected the data and NS extracted the data for analysis. AB and NS revised the manuscript for important intellectual content, approved the final version for submission, and agree to be accountable for all aspects of the work. All others who contributed to this document, but did not meet authorship criteria, are acknowledged.\n\n\nCompeting interests\n\n\n\nNS has received support for attendance of educational meetings by Gilead, Norgine and Bristol-Myers Squibb, and has received honoraria for educational presentations from Gilead and Bristol-Myers Squibb. AB has acted in an advisory capacity and/or received speaker honoraria from AbbVie, Bristol-Myers Squibb, Gilead, Janssen and Merck, and has been awarded grant money from Gilead to support academic research.\n\nUnder the direction of the authors, Amy MacLucas from iS LifeScience drafted the initial version of the manuscript and provided editorial support throughout its development. Editorial support was funded by Gilead Sciences.\n\n\nGrant information\n\nEditorial support for this manuscript was funded by Gilead Sciences.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to thank all the staff at the Liver and Antiviral Centre at St Mary’s Hospital, London, as well as all the patients who were involved in this cohort. We thank Biomedical Research Council for ongoing support of academic work at Imperial College Department of Hepatology. The authors would like to thank Amy MacLucas from iS LifeScience for her editorial support in the preparation of this manuscript.\n\n\nReferences\n\nWorld Health Organization: Fact Sheet N° 164: Hepatitis C. 2015. Reference Source\n\nLozano R, Naghavi M, Foreman K, et al.: Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380(9859): 2095–2128. PubMed Abstract | Publisher Full Text\n\nPublic Health England: Hepatitis C in the UK 2015 report. 2015. Reference Source\n\nSmith DB, Bukh J, Kuiken C, et al.: Expanded classification of hepatitis C virus into 7 genotypes and 67 subtypes: updated criteria and genotype assignment web resource. Hepatology. 2014; 59(1): 318–327. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManns MP, McHutchison JG, Gordon SC, et al.: Peginterferon alfa-2b plus ribavirin compared with interferon alfa-2b plus ribavirin for initial treatment of chronic hepatitis C: a randomised trial. Lancet. 2001; 358(9286): 958–965. PubMed Abstract | Publisher Full Text\n\nFried MW, Shiffman ML, Reddy KR, et al.: Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection. N Engl J Med. 2002; 347(13): 975–982. PubMed Abstract | Publisher Full Text\n\nHadziyannis SJ, Sette H Jr, Morgan TR, et al.: Peginterferon-alpha2a and ribavirin combination therapy in chronic hepatitis C: a randomized study of treatment duration and ribavirin dose. Ann Intern Med. 2004; 140(5): 346–355. PubMed Abstract | Publisher Full Text\n\nSulkowski MS, Cooper C, Hunyady B, et al.: Management of adverse effects of Peg-IFN and ribavirin therapy for hepatitis C. Nat Rev Gastroenterol Hepatol. 2011; 8(4): 212–223. PubMed Abstract | Publisher Full Text\n\nDudley T, O'Donnell K, Haydon G, et al.: Disappointing results of combination therapy for HCV? Gut. 2006; 55(9): 1362–1363. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamachandran P, Fraser A, Agarwal K, et al.: UK consensus guidelines for the use of the protease inhibitors boceprevir and telaprevir in genotype 1 chronic hepatitis C infected patients. Aliment Pharmacol Ther. 2012; 35(6): 647–662. PubMed Abstract | Publisher Full Text\n\nManns MP, von Hahn T: Novel therapies for hepatitis C - one pill fits all? Nat Rev Drug Discov. 2013; 12(8): 595–610. PubMed Abstract | Publisher Full Text\n\nZopf S, Kremer AE, Neurath MF, et al.: Advances in hepatitis C therapy: What is the current state - what come's next? World J Hepatol. 2016; 8(3): 139–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTaylor LE, Swan T, Matthews GV: Management of hepatitis C virus/HIV coinfection among people who use drugs in the era of direct-acting antiviral-based therapy. Clin Infect Dis. 2013; 57(Suppl 2): S118–124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDore G, Altice F, Litwin AH, et al.: C-EDGE CO-STAR: efficacy of grazoprevir and elbasvir in persons who inject drugs (PWID) receiving opioid agonist therapy. Hepatology. 2015; 62(Suppl 1): 227A. Publisher Full Text\n\nWorld Medical Association: World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013; 310(20): 2191–2194. PubMed Abstract | Publisher Full Text\n\nSelvapatt N, Habibi MS, Brown A: A European single centre experience of management of hepatitis C virus genotype 4 infection with pegylated-interferon and ribavirin. J Med Virol. 2015; 87(10): 1716–1721. PubMed Abstract | Publisher Full Text\n\nPublic Health England: Hepatitis C: commissioning template for estimated disease prevalence and treatment. 2014. Reference Source\n\nAfdhal N, Zeuzem S, Kwo P, et al.: Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection. N Engl J Med. 2014; 370(20): 1889–1898. PubMed Abstract | Publisher Full Text\n\nAfdhal N, Reddy KR, Nelson DR, et al.: Ledipasvir and sofosbuvir for previously treated HCV genotype 1 infection. N Engl J Med. 2014; 370(16): 1483–1493. PubMed Abstract | Publisher Full Text\n\nKowdley KV, Gordon SC, Reddy KR, et al.: Ledipasvir and sofosbuvir for 8 or 12 weeks for chronic HCV without cirrhosis. N Engl J Med. 2014; 370(20): 1879–1888. PubMed Abstract | Publisher Full Text\n\nFeld JJ, Kowdley KV, Coakley E, et al.: Treatment of HCV with ABT-450/r-ombitasvir and dasabuvir with ribavirin. N Engl J Med. 2014; 370(17): 1594–1603. PubMed Abstract | Publisher Full Text\n\nZeuzem S, Dusheiko GM, Salupere R, et al.: Sofosbuvir and ribavirin in HCV genotypes 2 and 3. N Engl J Med. 2014; 370(21): 1993–2001. PubMed Abstract | Publisher Full Text\n\nZeuzem S, Jacobson IM, Baykal T, et al.: Retreatment of HCV with ABT-450/r-ombitasvir and dasabuvir with ribavirin. N Engl J Med. 2014; 370(17): 1604–1614. PubMed Abstract | Publisher Full Text\n\nSulkowski MS, Gardiner DF, Rodriguez-Torres M, et al.: Daclatasvir plus sofosbuvir for previously treated or untreated chronic HCV infection. N Engl J Med. 2014; 370(3): 211–221. PubMed Abstract | Publisher Full Text\n\nFoster GR, Afdhal N, Roberts SK, et al.: Sofosbuvir and Velpatasvir for HCV Genotype 2 and 3 Infection. N Engl J Med. 2015; 373(27): 2608–2617. PubMed Abstract | Publisher Full Text\n\nSherman M, Yoshida EM, Deschenes M, et al.: Peginterferon alfa-2a (40KD) plus ribavirin in chronic hepatitis C patients who failed previous interferon therapy. Gut. 2006; 55(11): 1631–1638. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLagging M, Rembeck K, Rauning Buhl M, et al.: Retreatment with peg-interferon and ribavirin in patients with chronic hepatitis C virus genotype 2 or 3 infection with prior relapse. Scand J Gastroenterol. 2013; 48(7): 839–847. PubMed Abstract | Publisher Full Text\n\nManns MP, Wedemeyer H, Cornberg M: Treating viral hepatitis C: efficacy, side effects, and complications. Gut. 2006; 55(9): 1350–1359. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWitthoeft T, Hueppe D, John C, et al.: Efficacy and tolerability of peginterferon alfa-2a or alfa-2b plus ribavirin in the daily routine treatment of patients with chronic hepatitis C in Germany: the PRACTICE study. J Viral Hepat. 2010; 17(7): 459–468. PubMed Abstract | Publisher Full Text\n\nMcGowan CE, Fried MW: Barriers to hepatitis C treatment. Liver Int. 2012; 32(Suppl 1): 151–156. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcgowan CE, Monis A, Bacon BR, et al.: A global view of hepatitis C: physician knowledge, opinions, and perceived barriers to care. Hepatology. 2013; 57(4): 1325–1332. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYounossi ZM, Stepanova M, Nader F, et al.: The patient’s journey with chronic hepatitis C from interferon plus ribavirin to interferon- and ribavirin-free regimens: a study of health-related quality of life. Aliment Pharmacol Ther. 2015; 42(3): 286–295. PubMed Abstract | Publisher Full Text\n\nSpiegel BM, Younossi ZM, Hays RD, et al.: Impact of hepatitis C on health related quality of life: a systematic review and quantitative assessment. Hepatology. 2005; 41(4): 790–800. PubMed Abstract | Publisher Full Text\n\nDan AA, Kallman JB, Srivastava R, et al.: Impact of chronic liver disease and cirrhosis on health utilities using SF-6D and the health utility index. Liver Transpl. 2008; 14(3): 321–326. PubMed Abstract | Publisher Full Text\n\nBonkovsky HL, Snow KK, Malet PF, et al.: Health-related quality of life in patients with chronic hepatitis C and advanced fibrosis. J Hepatol. 2007; 46(3): 420–431. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYounossi ZM, Stepanova M, Henry L, et al.: Effects of sofosbuvir-based treatment, with and without interferon, on outcome and productivity of patients with chronic hepatitis C. Clin Gastroenterol Hepatol. 2014; 12(8): 1349–59.e13. PubMed Abstract | Publisher Full Text\n\nTalal AH, LaFleur J, Hoop R, et al.: Absolute and relative contraindications to pegylated-interferon or ribavirin in the US general patient population with chronic hepatitis C: results from a US database of over 45 000 HCV-infected, evaluated patients. Aliment Pharmacol Ther. 2013; 37(4): 473–481. PubMed Abstract | Publisher Full Text\n\nYounossi ZM, Jiang Y, Smith NJ, et al.: Ledipasvir/sofosbuvir regimens for chronic hepatitis C infection: Insights from a work productivity economic model from the United States. Hepatology. 2015; 61(5): 1471–1478. PubMed Abstract | Publisher Full Text\n\nSimmons B, Saleem J, Heath K, et al.: Long-Term Treatment Outcomes of Patients Infected With Hepatitis C Virus: A Systematic Review and Meta-analysis of the Survival Benefit of Achieving a Sustained Virological Response. Clin Infect Dis. 2015; 61(5): 730–740. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShoeb D, Rowe IA, Freshwater D, et al.: Response to antiviral therapy in patients with genotype 3 chronic hepatitis C: fibrosis but not race encourages relapse. Eur J Gastroenterol Hepatol. 2011; 23(9): 747–753. PubMed Abstract | Publisher Full Text\n\nCharlton M, Gane E, Manns MP, et al.: Sofosbuvir and ribavirin for treatment of compensated recurrent hepatitis C virus infection after liver transplantation. Gastroenterology. 2015; 148(1): 108–117. PubMed Abstract | Publisher Full Text\n\nFoster GR, Irving WL, Cheung MC, et al.: Impact of direct acting antiviral therapy in patients with chronic hepatitis C and decompensated cirrhosis. J Hepatol. 2016; 64(6): 1224–1231. PubMed Abstract | Publisher Full Text\n\nSelvapatt N, Brown A: Dataset 1 in: A real-world intention-to-treat analysis of a decade’s experience of treatment of hepatitis C with interferon-based therapies. F1000Research. 2016. Data Source"
}
|
[
{
"id": "16119",
"date": "06 Sep 2016",
"name": "Mark Wright",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article stands as a final word on a closed era of HCV treatment.\nI wonder if in the rapidly changing world of HCV treatments the introduction should be updated a little, especially the second to last paragraph that implies that in the real world some patients with G1 might still be treated with PEG RIBA.\nI also question the relavence of the presenting of the data separating referrals to the anti viral unit (as a sort of ITT analysis) and those actually treated (per protocol esque analysis). This is a little artificial as we don't know how many people were seen in other hepatology clinics and never referred because they were either deemed unsuitable or uninterested.\nOther than these few tweaks I approve.",
"responses": []
},
{
"id": "16493",
"date": "22 Sep 2016",
"name": "Andrew Ustianowski",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMany thanks for asking me to review this article.\n\nThough no longer topical (as the treatments have progressed) this paper remains of interest as it provides some real-world insights, which are unlikely to be repeated, from an era of HCV care.\n\nI believe it is suitable for indexing, but have some minor recommendations.\n\nI wonder if it is overly long in its present form, with some minor repetition in introduction & discussion, and with topics discussed that are not directly relevant to the questions posed. However this will ultimately depend on the editorial flexibility and approach of the journal.\n\nAfter reading the abstract I was especially interested in the data on those that did not receive treatment (which is largely missing from the literature), but on reading the article itself it is apparent that this data is not of such utility unfortunately. This is because there are confounders in that other specialist clinicians have already effectively excluded many individuals by not referring on for treatment. There is also the issue of not considering those that did not attend. Though unavoidable issues with the design of this research, it does relate to the issue of using ITT analyses (as the denominators used are not translatable to any other settings). It also means that comments such as ‘a greater proportion of patients received treatment compared with those who did not receive treatment (84.4% vs15.6%, respectively)’ do not provide any useful information (as these individuals have effectively been pre-screened as being more suitable for therapy).\nThe study examined only G1-3 and excluded G4-6. It is not clear as to why this was the case and also it means that statements such as ‘In line with national estimates published by Public Health England, over 90% of the patients in this study had HCV genotype 1 or 3 infections’ should be avoided as you are not comparing like with like.\nThe authors comment that their SVR rates for G1 and G2 are comparable to the trial data, potentially due to the clinical and patient support infrastructure they have in place – but that infrastructure was presumably also there for the G3 patients (who had a lower than expected SVR rate) and therefore I think such conclusions may be too speculative.\n\nOther comments:\n\nIt is more generally accepted that there are 6 distinct genotypes of HCV – not 7 as stated in paragraph 2 of the introduction.\n\nThe graphing in figures 4 and 5 does not need both columns as the outcome for each is dichotomous (either SVR or failure) – therefore just the SVR column would be better for clarity.\n\nI could not link on my computer to the datasets and so could not assess them.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2061
|
https://f1000research.com/articles/5-1005/v1
|
26 May 16
|
{
"type": "Research Article",
"title": "Identification of selective inhibitors of RET and comparison with current clinical candidates through development and validation of a robust screening cascade",
"authors": [
"Amanda J. Watson",
"Gemma V. Hopkins",
"Samantha Hitchin",
"Habiba Begum",
"Stuart Jones",
"Allan Jordan",
"Sarah Holt",
"H. Nikki March",
"Rebecca Newton",
"Helen Small",
"Alex Stowell",
"Ian D. Waddell",
"Bohdan Waszkowycz",
"Donald J. Ogilvie",
"Gemma V. Hopkins",
"Samantha Hitchin",
"Habiba Begum",
"Stuart Jones",
"Allan Jordan",
"Sarah Holt",
"H. Nikki March",
"Rebecca Newton",
"Helen Small",
"Alex Stowell",
"Ian D. Waddell",
"Bohdan Waszkowycz",
"Donald J. Ogilvie"
],
"abstract": "RET (REarranged during Transfection) is a receptor tyrosine kinase, which plays pivotal roles in regulating cell survival, differentiation, proliferation, migration and chemotaxis. Activation of RET is a mechanism of oncogenesis in medullary thyroid carcinomas where both germline and sporadic activating somatic mutations are prevalent.\n\nAt present, there are no known specific RET inhibitors in clinical development, although many potent inhibitors of RET have been opportunistically identified through selectivity profiling of compounds initially designed to target other tyrosine kinases. Vandetanib and cabozantinib, both multi-kinase inhibitors with RET activity, are approved for use in medullary thyroid carcinoma, but additional pharmacological activities, most notably inhibition of vascular endothelial growth factor - VEGFR2 (KDR), lead to dose-limiting toxicity. The recent identification of RET fusions present in ~1% of lung adenocarcinoma patients has renewed interest in the identification and development of more selective RET inhibitors lacking the toxicities associated with the current treatments.\n\nIn an earlier publication [Newton et al, 2016; 1] we reported the discovery of a series of 2-substituted phenol quinazolines as potent and selective RET kinase inhibitors. Here we describe the development of the robust screening cascade which allowed the identification and advancement of this chemical series. Furthermore we have profiled a panel of RET-active clinical compounds both to validate the cascade and to confirm that none display a RET-selective target profile.",
"keywords": [
"RET",
"KDR",
"Lung adenocarcinoma",
"Screening cascade",
"Selectivity"
],
"content": "Introduction\n\nRET is a receptor tyrosine kinase (TK) expressed primarily on derived neural crest and urogenital cells during embryonic development. It is required for maturation of several cell lineages of the peripheral nervous system, kidney morphogenesis and spermatogenesis2. The glial derived neurotrophic factor (GDNF) family of ligands bind RET in association with one of four glycosyl phosphatidylinositol (GPI) anchored GDNF family α-receptors (GFRα), triggering RET dimerization, followed by auto-phosphorylation of specific tyrosine residues within the C-terminal chain and trans-phosphorylation of intracellular signalling cascades. These downstream signalling networks play a key role in regulating cell survival, differentiation, proliferation, migration and chemotaxis3.\n\nActivating mutations in RET (e.g. C634W and M918T) have been identified in familial and sporadic forms of medullary thyroid carcinomas (MTC4,5) and are associated with aggressive disease progression6. More recently, several groups independently identified RET rearrangements in 1–2% of lung adenocarcinoma (LAD) cases7–10. The RET fusion genes discovered in these studies include CCDC6-RET (already known as RET/PTC1 in papillary thyroid carcinoma) as well as a novel fusion with KIF5B (kinesin family member 5B), encoding a coiled coil domain, generated by pericentric inversion in chromosome 10. The coiled-coil domains present in the fusion partner promote overexpression and ligand-independent dimerization leading to constitutive activation of RET. These studies also demonstrated that the resulting fusion proteins are oncogenic, and that their inhibition has therapeutic potential. Importantly, the RET fusions are mutually exclusive with other known drivers in LAD (e.g. KRAS, endothelial growth factor (EGFR), EML4-anaplastic lymphoma kinase (ALK)), further supporting a role for RET as a unique driver of malignancy in these tumors. RET-positive patients represent a well-defined population with specific features: all are adenocarcinomas, and patients tend to be non-smokers and to be younger than the median age for lung cancer patients11.\n\nAt present, there are no known specific RET inhibitors in clinical development, although many potent inhibitors of RET have been opportunistically identified through selectivity profiling of compounds initially designed to target other TKs. The small molecule inhibitors vandetanib and cabozantinib are perhaps the best examples of such compounds. Although both have been approved for the treatment of advanced metastatic MTC12,13, RET inhibition is a secondary pharmacology of these drugs, which were initially developed as inhibitors of other TKs. Both agents target KDR, whilst vandetanib has additional activity versus EGFR and cabozantinib versus MET. These compounds are now under investigation for the treatment of RET fusion positive LAD. A preliminary report of a phase II trial14 of cabozantinib confirmed partial responses in two of three RET-positive patients11; the third patient presented with prolonged stable disease. The activity of vandetanib in RET fusion positive patients has been demonstrated in two case reports15. However significant toxicity (e.g. rash, diarrhoea, hypertension) resulting from inhibition of off-target kinases, particularly KDR, has led to dose reductions in clinical trials (11–13) and is likely to compromise the use of both these agents in clinical settings16. Thus, there is a clear need for selective RET inhibitors which do not display the non-pharmacological toxicities associated with the current treatments and enable more potent and sustained inhibition of RET signalling. These agents may offer greater clinical benefit for patients with RET mutant cancers and widen the scope for the clinical use of RET inhibitors17.\n\nThe role of RET in this subset of LAD has heightened interest in re-purposing a number of other clinically approved inhibitors, shown to have RET activity in pre-clinical studies. Sunitinib, sorafenib, ponatinib and lenvatinib, all multi-kinase TK inhibitors with some RET activity, are currently under investigation in numerous phase II clinical trials14 for treatment of RET fusion positive LAD18. Sunitinib, already approved for the treatment of imatinib-resistant gastrointestinal stromal tumors (GIST), advanced renal carcinoma and advanced pancreatic neuroendocrine tumors, is the subject of a phase II study in certain types of LAD tumors, including those harbouring a RET fusion. Sorafenib, also approved for several indications including kidney and liver cancer, has demonstrated preclinical activity in RET models but has yet to be tested in patients selected based on RET fusion status. Some efficacy in advanced MTC has been reported for lenvatinib19, however tumor response did not correlate with RET mutation status and the observed toxicity profile was consistent with KDR inhibition. A phase II study of lenvatinib in RET fusion positive LAD is ongoing14. Ponatinib is also a multi-targeted, broad-spectrum tyrosine kinase inhibitor20, approved in late 2012 for patients with resistant or intolerant chronic myeloid leukemia and Philadelphia chromosome-positive acute lymphoblastic leukemia. Ponatinib was withdrawn shortly afterwards due to serious safety concerns but was later returned to the market with additional warnings in the product information. An investigational phase II clinical trial of ponatinib in LAD patients selected based on RET mutation status is currently ongoing14. Alectinib is a highly selective ALK inhibitor (median inhibitory concentration of 1.9 nM for ALK activity), recently approved by the FDA for the treatment of patients with ALK positive LAD who progressed on crizotinib. Preclinical data demonstrating activity of alectinib in RET mouse models21 has led to its investigation as a treatment for RET fusion positive LAD as part of the DARWIN II trial14. The compounds described above are all classic TK inhibitors but in 2010, Alfano et al.22 proposed an alternative approach for targeting RET in LAD involving inhibition of HSP90 (heat shock protein 90kDa). HSP90 is a molecular chaperone that plays a central role in regulating the correct folding, stability and function of numerous proteins23. Inhibition of HSP90 activity results in aggregation or proteasomal degradation of these proteins. RET, along with other driver kinases such as EGFR and ALK is a HSP90 client protein and as such requires HSP90 for protein stability and function. Thus, targeting the chaperone function of HSP90 offers an alternative to direct kinase inhibition for therapeutic intervention in RET driven cancer. Clinical evaluation of this approach is currently being assessed as part of a larger study in stage IV LAD patients with driver molecular alterations other than EGFR mutations14.\n\nIn an earlier publication1 we reported the discovery of 2-substituted phenol quinazolines as potent RET kinase inhibitors with improved KDR selectivity; here we describe development of the robust screening cascade which allowed us to achieve that goal and additionally profile existing clinical compounds with RET activity in order to assess them against our compound target profile.\n\n\nMaterials and methods\n\nClinical compounds were purchased from SelleckChem. Paterson Drug Discovery (PDD) compounds were synthesised in-house by methods described in an earlier publication1. All compounds were dissolved at 20mM in dimethyl sulfoxide (DMSO, Sigma) and stored at -20°C or in a desiccator at room temperature.\n\nRET and KDR kinase activities were measured according to methods previously described using the HTRF kinEASE kit (CisBio)1. For measurement of mutant RET (M918T; Carna Bioscience), the following modifications were made: 50pM RET (M918T), 9µM ATP and 1µM substrate. For the slow binding studies, the standard kinase assay procedure was used but with varying time of compound pre-incubation with RET prior to the addition of ATP. Reversibility of binding was determined by measuring the recovery of enzymatic activity after a rapid and large dilution of the enzyme-inhibitor complex. RET enzyme was incubated at 100 fold the concentration normally required for the standard screening assay, with a concentration of inhibitor equivalent to 10 fold the IC50. After 15 minutes of incubation, this mixture was diluted 100 fold in the reaction buffer containing the enzyme substrates to start the reaction. This diluted the enzyme to the standard assay concentration and the compound from 10 fold to 0.1 fold of the IC50 concentration. Theoretically, following this 100-fold dilution of the pre-incubation mixture, 91% enzyme activity will be recovered for a fully reversible inhibitor. However, taking into account assay variability, compounds showing >75% enzymatic activity recovery compared to the no inhibitor positive control were classed as fully reversible, those with <75% but where recovery of some activity was clearly detected were classed as slowly reversible (i.e. would eventually reach >75% recovery of activity). To test whether compounds were ATP competitive, assays were performed under standard conditions with 9µM ATP (Km), and then repeated at 450µM ATP (50× Km). By increasing the amount of ATP in the reaction by 50 fold, the IC50 value should increase for competitive compounds, theoretically by a ratio of 25. For non-competitive compounds the ratio should be 0.5.\n\nMZ-CRC-1 and LC-2/ad cells were cultured in advanced DMEM/F12 media (Invitrogen), supplemented with 5% fetal bovine serum (FBS, Invitrogen) and 2mM Glutamax (Invitrogen) and incubated at 37°C in 5% CO2/air. The BaF3 cell lines, expressing KIF5B-RET (gift from Pasi Janne10) and KDR (Advanced Cellular Dynamics, San Diego) were maintained in RPMI-1640 (Invitrogen) media containing 10% Hyclone FBS (Scientific Laboratory Supplies) with the addition of 1µg/mL puromycin (Sigma) for the KDR cell line. Non-modified BaF3 cells (WT; DSMZ, Germany) were maintained in RPMI-1640 media (Invitrogen) containing 10% FBS (Invitrogen) and supplemented with 10 ng/mL recombinant mouse IL-3 (R&D systems).\n\nThe RET proof of mechanism assay (POM) assay measures the compound effect on the target. Active forms of RET and KDR are phosphorylated and thus compound inhibitory activity may be measured directly by quantifying levels of the phosphoproteins, pRET and pKDR, remaining after cell treatment. MZ-CRC-1 cells harbour the M918T mutation and thus express constitutively high levels of pRET and although pKDR is barely detectable it is possible to increase expression with ligand stimulation thus allowing measurement of RET and KDR inactivation within the same cell lysate. MZ-CRC-1 cells were seeded into 96-well plates at 100,000 cells per well in 100µL culture medium. After 24 hours, medium was replaced with 100µl of serum-free medium and cells were incubated overnight. Compounds were dispensed into the appropriate wells using an acoustic liquid handling platform (LABCYTE). Cells were incubated with compound for 2h, followed by vascular endothelial growth factor (human VEGF 165, R&D Systems) ligand stimulation (50ng/ml for 5 minutes at 37°C). VEGFR treatment does not affect levels of pRET (data not shown). Cells were washed with 100µL ice-cold PBS and 30µL lysis buffer (Cell Signalling Technology) added. Plates were incubated at 4°C for 1 hour, resulting lysates were transferred to another 96-well plate and protein concentration determined using the Millipore Direct Detect infrared spectrometer. Levels of pRET and pKDR in the cell lysates were determined, according to the manufacturer’s instructions, using the pRET (panTyr) and pVEGFR-2 (Tyr1175) PathScan sandwich ELISA kits (Cell Signalling Technology). Compound IC50s were based on levels of phosphoprotein normalised to control values. Selectivity (versus KDR) was calculated by dividing KDR IC50 value by RET IC50 value.\n\nCompound effects on pRET levels in LC-2/ad cells were measured as described above except that 30,000 cells per well were plated and the ligand stimulation step was omitted; KDR protein is not detectable in this cell line and therefore it is not possible to determine compound effects on levels of pKDR.\n\nThe RET proof of principle (POP) assay measures compound effects on cell proliferation in a disease relevant model. For routine screening we compared anti-proliferative effects of the compounds in the disease model, MZ-CRC-1 (RET (M918T)) versus a control, non-RET expressing cell line, Hek293. This allows us to demonstrate translation of mechanistic effects measured in the POM assay into RET-specific phenotypic effects in cells. MZ-CRC-1 and Hek293 (control, no RET expression) cells were seeded into 384 plates at 4000 and 1000 cells per well respectively in 30µL culture medium and confluence monitored at 4 hourly intervals using the IncuCyte ZOOM live cell imaging platform (Essen). After 48 hours, compounds were dispensed as described above and cells incubated until confluence of control cells reached 80–90%. Compound IC50s were calculated based on cell confluency at this time point normalised to control values. Non-specific toxicity margin was calculated by dividing IC50 value obtained for the Hek293 control cells by that for the MZ-CRC-1 cells.\n\nThe POP assay has also been used to measure compound effects in other disease relevant cell lines, for example, LC-2/ad, an LAD cell line harbouring the CCDC6-RET fusion. LC-2/ad cells were seeded into 96 well plates at 4000 cells per well in 100µL serum free culture medium and incubated at 37°C in 5% CO2. After 48 hours compounds were diluted to 2x final concentration and added to appropriate wells in 100µL serum-free culture medium. Once control cells had reached 80–90% confluence, protein content was measured using the sulforhodamine B (SRB) assay24. Compound IC50s were based on protein content (proportional to cell number) normalised to control values.\n\nThis assay was performed as described previously1. Selectivity values were calculated as described above.\n\nFollowing compound treatment (1µM, 24 hours), MZ-CRC-1 cell lysates (50 μg) were subjected to polyacrylamide gel electrophoresis (PAGE) and semi-dry transfer to nitrocellulose membrane using the Bio-Rad Trans-Blot Turbo system and Trans-Blot Turbo transfer packs. Membranes were blocked overnight at 4°C in phosphate-buffered saline containing 0.1% Tween (PBST; Sigma) and 5% non-fat dried milk powder (marvel). Primary and secondary antibody incubations were performed at room temperature in PBST/0.5% marvel with 3× PBST washes post incubation. pRET (Santa Cruz #SC-20252-R) and total RET (Santa Cruz #sc-167) antibodies were used at 1:500 dilutions; GAPDH (Cell Signaling Technology #2118L) and secondary goat anti rabbit IgG HRP-linked (Cell Signaling Technology #7074) antibodies were used at 1:1000. Proteins were visualized by chemiluminescent detection of peroxidase activity using SuperSignal reagent (Pierce), and images were captured using the Syngene imager and GeneSys software.\n\n\nResults and discussion\n\nOur aim was to develop and validate a robust screening cascade to support the identification and development of potent and selective RET inhibitors using vandetanib as the starting point1 and to assess these attributes in clinical compounds reported to have RET activity. All compounds were initially assayed biochemically for activity versus RET and KDR enzyme. Once potency and selectivity had been confirmed, and structure activity relationships (SAR) demonstrated, for a number of compounds within the anilinoquinazoline series1, we performed further biochemical studies to investigate the mechanism of RET inhibition using selected compounds including our starting point, vandetanib (Figure 1). Some inhibitors bind to, or dissociate from the target enzyme slowly, leading to time dependent inhibition25. Failure to identify this can lead to an underestimation of biochemical potency and misleading SAR. To investigate this, compounds were assayed under standard conditions but pre-incubated with RET for between 0 and 60 minutes prior to addition of ATP. Pre-incubation did not significantly affect IC50 values indicating that the anilinoquinazolines are not slow binders and that a 15 minutes pre-incubation, as used in our standard assay, is sufficient to allow the reaction to reach equilibrium (Figure 1A). Although diverse in primary amino acid sequence, the human kinases share a great degree of similarity in their 3D structures, especially in their catalytically active kinase domain where the ATP-binding pocket is located. Kinase inhibitors can be grouped into two classes, based on binding mode: irreversible and reversible. The former tend to bind covalently with a reactive nucleophilic cysteine residue proximal to the ATP-binding site, resulting in the blockage of the ATP site and irreversible inhibition. Reversible inhibitors can be further classified into four main types, competitive, non-competitive, uncompetitive and mixed inhibition26. Our data (Figure 1B) indicates that the anilinoquinazolines under investigation here fall into the reversible, ATP competitive category which is not surprising since it has been demonstrated previously that vandetanib, a related compound and our starting point, exhibits this mode of binding27.\n\n(A) Effect of increasing pre-incubation time (5–60 mins) on compound IC50 value. (B) Table showing results of reversibility and ATP competition studies.\n\nCompounds meeting defined criteria for potency and selectivity in the biochemical assay were then assessed in a cellular POM assay. Initially we developed an ELISA-based POM assay, measuring changes in levels of (active) phosphorylated RET (M918T) and KDR, within the same MTC cell line (MZ-CRC-1).\n\nHowever, it soon became apparent that compared to KDR there was a poor correlation between RET biochemical and cell-based IC50 measurements (Figures 2A & B). In addition, the drop off in potency from the biochemical to cellular assay was generally much greater for RET than KDR, in effect compressing the selectivity margins, often from >100-fold in the biochemical assay to parity (or worse) in the cellular assay. Compound permeability was not believed to be the cause of the disparity as the magnitude of the drop off was not the same for both RET and KDR, despite measurement of both proteins in the same cell lysate. A previous report28 had indicated that a single mutation within the kinase domain of RET could alter compound inhibitory activity. In order to address the possibility that some of the discrepancy could be due to our measurement of WT protein activity biochemically but mutant RET (M918T) protein in the MZ-CRC-1 cells, we compared the biochemical activity of selected compounds versus both RET (WT) and RET (M918T) protein. The compound IC50 data for the two proteins correlate well (Figure 2C) and therefore indicate that the mutation does not significantly affect biochemical activity for this class of compound. In order to follow this up in the cellular context we extended our repertoire of assays to enable robust measurement of RET activity in TT29, LC-2/ad30,31 and mouse BaF310 cells harbouring a C634W mutation, a CCDC6-RET fusion and a KIF5B-RET fusion respectively. We assessed relative RET potencies for a number of tool compounds (including vandetanib and cabozantinib) across these cell lines and found that activities correlated well (Supplementary Figure 1), further supporting the notion that this class of compounds are equally active versus mutant and WT forms of the protein.\n\nIn order to demonstrate POP in a disease relevant cell line, we developed and validated a proliferation endpoint assay in the MTC line MZ-CRC-1; non-specific toxicity was evaluated by measuring the same endpoint in Hek293, a human embryonic kidney line which does not express RET. Compounds exhibiting RET potency (<500nM) and selectivity (>10x) in the POM assay were selected for POP assay screening. Our data show that there is a good correlation between the POM and POP IC50 values measured in the MZ-CRC-1 (Figure 3A) cells and other RET models (e.g. LC-2/ad, Supplementary Figure 2) indicating that the RET inhibitory activity that we observe in our POM assay translates into anti-proliferative effects in our MTC and LAD disease models. In addition, we have assayed a selection of representative compounds for anti-proliferative effects in LC-2/ad and as for the POM assay, data correlate well with that obtained in MZ-CRC-1 (Figure 3B). Therefore, we are confident that activities observed in our routine MZ-CRC-1 POM and POP screens are predictive for the other cell line models.\n\nA) RET enzyme vs RET cellular POM B) KDR enzyme vs KDR cellular POM C) RET enzyme WT vs RET (M918T).\n\nCorrelation of cellular data: A) POM vs POP IC50 data in MZ-CRC-1 B) POP vs POP IC50 data in MZ-CRC-1 and LC-2/ad C) POM vs POM IC50 data in MZ-CRC-1 and KIF5B-RET (BaF3) D) POM vs POM selectivity data in MZ-CRC-1 and KIF5B-RET (BaF3).\n\nAlthough at this stage we were confident that we could drive our internal chemistry effort based on the cellular data, it was clear that performance in the biochemical assay was not always a good predictor of RET cellular activity and therefore should not be used as a pre-screen for this chemical series. To address this issue this we developed and validated a higher throughput POM assay allowing parallel assessment of all compounds in both cellular and biochemical assays. This BaF3 proliferation assay platform, employing cell lines dependent upon RET or KDR for survival alongside WT IL-3 dependent control cells, is robust with 4x higher throughput than the original MZ-CRC-1 POM ELISA. More importantly the proliferation IC50 for both RET (Figure 3C) and KDR in the BaF3 models correlate well with those generated using the respective phosphorylation endpoint assays thus maintaining the selectivity values previously observed (Figure 3D). Accordingly, we introduced this as the routine POM screening assay with the option to further evaluate compound effects on phosphoprotein levels using the ELISA POM if necessary.\n\nChemical optimisation of the series ultimately led to the identification of a number of potent and selective RET inhibitors (e.g. PDD16860 and PDD16964; Table 1). Clinical RET compounds currently under investigation in LAD were also profiled using the screening cascade (Table 1). With the possible exception of alectinib, which it was not possible to dose at the highest concentrations due to solubility issues, none of the kinase inhibitors tested met our criteria for selectivity versus KDR (i.e. >10x). Several reports in the literature21,32, which we have subsequently confirmed in-house (data not shown), demonstrate that alectinib is a much better cellular inhibitor of ALK (IC50 in H2228 cells = 10nM) than RET (IC50 in MZ-CRC-1 cells = 80nM). Given this, along with its limited solubility and required dosing of 300–600mg bid for clinical efficacy in ALK driven LAD, therapeutic inhibition of RET is unlikely at clinically relevant doses. Although not a direct inhibitor of RET activity, we have also profiled ganetespib, a HSP90 inhibitor. Ganetespib did reduce levels of active RET, through protein degradation (Supplementary Figure 3) but was not RET-selective and exhibited more toxicity (toxicity margin = 0.7x) in the control non-RET driven Hek293 cells compared to the MTC RET-driven model, MZ-CRC-1. This indicates that the pleiotropic effects of HSP90 inhibition on numerous client proteins (including KDR) is likely to elicit non-specific toxicity at therapeutically active doses. Other aspects of the screening cascade (Figure 4) are already in place and as discussed in an earlier publication1, selected compounds have been assessed in in vitro DMPK assays as part of the chemical optimisation. Furthermore, in vivo models and capabilities (Figure 4) have been established and validated (data not shown) and several tool compounds are currently undergoing assessment in PK/PD and efficacy studies. To date, our robust screening workflows have proved very successful in identifying potent, selective tool compounds and will continue to support our active pursuit of clinical candidates suitable for evaluation in RET fusion positive LAD patients.\n\nThe desired profile values indicate our target values for selectivity and toxicity margin. All values represent the geometric mean of at least four independent measurements. Figures are highlighted depending upon whether they easily (green); clearly do not (red) or almost (orange) meet the target criteria. PDD16860 and PDD16964 are examples of QZ compounds developed by screening through the cascade.\n\n1Solubility of alectinib is limited, maximum dose achieved=1µM.\n\n2Ganetespib is not a kinase inhibitor and therefore it is not possible to measure biochemical activity.\n\nSol = solubility, Stab = stability, Perm = permeability, BB = Blood-binding, Cyps = Cytochrome P450s.\n\n\nConclusions\n\nWe have established a robust screening cascade to enable the identification and development of a clinical candidate compound demonstrating potent and selective RET activity. Furthermore we have profiled clinical compounds currently under investigation for treatment of LAD confirming that they do not meet our target profile. We are continuing our efforts to develop a clinical candidate and will report on further progress in a future publication.\n\n\nData availability\n\nF1000Research: Dataset 1. Datasets: selective inhibitors of RET and comparison with current clinical candidates through development and validation of a robust screening cascade, 10.5256/f1000research.8724.d12228033.",
"appendix": "Author contributions\n\n\n\nAJW, GH, SH, HB, SH, HNM designed and conducted the biological experiments, AJW, HS and AS conceptualized the experiments, SJ, RN and BW were responsible for chemical design and synthesis and AJ, IW and DO provided strategic direction. AJW prepared the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nThere are no competing financial interests to declare.\n\n\nGrant information\n\nThis work was supported by 6th Element Capital’s Pioneer fund and by Cancer Research UK (Grant C480/A11411 and C309/A8274).\n\n\nAcknowledgements\n\nWe are grateful to Pasi Janne for the kind gift of the BaF3 KIF5B-RET cell line and the other members of the Cancer Research UK Manchester Institute Drug Discovery Unit for valuable advice and support.\n\n\nSupplementary material\n\nIC50 data for MZ-CRC-1, LC-2/ad, TT and KIF5B-RET (BaF3) cells.\n\nWestern blot showing effects of compound treatment (1µM, 24 hours) on pRET, total RET and GAPDH levels in MZ-CRC-1 cells.\n\n\nReferences\n\nNewton R, Bowler KA, Burns EM, et al.: The discovery of 2-substituted phenol quinazolines as potent RET kinase inhibitors with improved KDR selectivity. Eur J Med Chem. 2016; 112: 20–32. PubMed Abstract | Publisher Full Text\n\nArighi E, Borrello MG, Sariola H: RET tyrosine kinase signaling in development and cancer. Cytokine Growth Factor Rev. 2005; 16(4–5): 441–467. PubMed Abstract | Publisher Full Text\n\nSantoro M, Carlomagno F, Melillo RM, et al.: Dysfunction of the RET receptor in human cancer. Cell Mol Life Sci. 2004; 61(23): 2954–2964. PubMed Abstract | Publisher Full Text\n\nSantoro M, Carlomagno F: Drug insight: Small-molecule inhibitors of protein kinases in the treatment of thyroid cancer. Nat Clin Pract Endocrinol Metab. 2006; 2(1): 42–52. PubMed Abstract | Publisher Full Text\n\nWells SA Jr, Santoro M: Targeting the RET pathway in thyroid cancer. Clin Cancer Res. 2009; 15(23): 7119–23. PubMed Abstract | Publisher Full Text\n\nElisei R, Cosci B, Romei C, et al.: Prognostic significance of somatic RET oncogene mutations in sporadic medullary thyroid cancer: a 10-year follow-up study. J Clin Endocrinol Metab. 2008; 93(3): 682–687. PubMed Abstract | Publisher Full Text\n\nJu YS, Lee WC, Shin JY, et al.: A transforming KIF5B and RET gene fusion in lung adenocarcinoma revealed from whole-genome and transcriptome sequencing. Genome Res. 2012; 22(3): 436–45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKohno T, Ichikawa H, Totoki Y, et al.: KIF5B-RET fusions in lung adenocarcinoma. Nat Med. 2012; 18(3): 375–377. PubMed Abstract | Publisher Full Text\n\nTakeuchi K, Soda M, Togashi Y, et al.: RET, ROS1 and ALK fusions in lung cancer. Nat Med. 2012; 18(3): 378–381. PubMed Abstract | Publisher Full Text\n\nLipson D, Capelletti M, Yelensky R, et al.: Identification of new ALK and RET gene fusions from colorectal and lung cancer biopsies. Nat Med. 2012; 18(3): 382–384. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDrilon A, Wang L, Hasanovic A, et al.: Response to Cabozantinib in patients with RET fusion-positive lung adenocarcinomas. Cancer Discov. 2013; 3(6): 630–635. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWells SA Jr, Robinson BG, Gagel RF, et al.: Vandetanib in patients with locally advanced or metastatic medullary thyroid cancer: a randomized, double-blind phase III trial. J Clin Oncol. 2012; 30(2): 134–141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElisei R, Schlumberger MJ, Müller SP, et al.: Cabozantinib in progressive medullary thyroid cancer. J Clin Oncol. 2013; 31(29): 3639–3646. PubMed Abstract | Publisher Full Text | Free Full Text\n\nhttps://clinicaltrials.gov/ (accessed February 2016).\n\nGautschi O, Zander T, Keller FA, et al.: A patient with lung adenocarcinoma and RET fusion treated with vandetanib. J Thorac Oncol. 2013; 8(5): e43–44. PubMed Abstract | Publisher Full Text\n\nSuguwara M, Ly T, Hershman JM: Medullary thyroid cancer--current treatment strategy, novel therapies and perspectives for the future. Horm Cancer. 2012; 3(5–6): 218–226. PubMed Abstract | Publisher Full Text\n\nMulligan LM: RET revisited: expanding the oncogenic portfolio. Nat Rev Cancer. 2014; 14(3): 173–186. PubMed Abstract | Publisher Full Text\n\nSong M: Progress in Discovery of KIF5B-RET Kinase Inhibitors for the Treatment of Non-Small-Cell Lung Cancer. J Med Chem. 2015; 58(9): 3672–3681. PubMed Abstract | Publisher Full Text\n\nSchlumberger M, Jarzab B, Cabanillas ME, et al.: A Phase II Trial of the Multitargeted Tyrosine Kinase Inhibitor Lenvatinib (E7080) in Advanced Medullary Thyroid Cancer. Clin Cancer Res. 2016; 22(1): 44–53. PubMed Abstract | Publisher Full Text\n\nDe Falco V, Buonocore P, Muthu M, et al.: Ponatinib (AP24534) is a novel potent inhibitor of oncogenic RET mutants associated with thyroid cancer. J Clin Enocrinol Metab. 2013; 98(5): E811–E819. PubMed Abstract | Publisher Full Text\n\nKodama T, Tsukaguchi T, Satoh Y, et al.: Alectinib shows potent antitumor activity against RET-rearranged non-small cell lung cancer. Mol Cancer Ther. 2014; 13(12): 2910–2918. PubMed Abstract | Publisher Full Text\n\nAlfano L, Guida T, Provitero L, et al.: RET is a heat shock protein 90 (HSP90) client protein and is knocked down upon HSP90 pharmacological block. J Clin Endocrinol Metab. 2010; 95(7): 3552–3557. PubMed Abstract | Publisher Full Text\n\nTaipale M, Jarosz DF, Lindquist S: HSP90 at the hub of protein homeostasis: emerging mechanistic insights. Nat Rev Mol Cell Biol. 2010; 11(7): 515–528. PubMed Abstract | Publisher Full Text\n\nVichai V, Kirtikara K: Sulforhodamine B colorimetric assay for cytotoxicity screening. Nat Protoc. 2006; 1(3): 1112–1116. PubMed Abstract | Publisher Full Text\n\nRudolph J, Xiao Y, Pardi A, et al.: slow inhibition and conformation selective properties of extracellular signal-regulated kinase 1 and 2 inhibitors. Biochemistry. 2015; 54(1): 22–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu P, Nielsen TE, Clausen MH: FDA-approved small-molecule kinase inhibitors. Trends Pharmacol Sci. 2015; 36(7): 422–39. PubMed Abstract | Publisher Full Text\n\nKnowles PP, Murray-Rust J, Kjaer S, et al.: Structure and chemical inhibition of the RET tyrosine kinase domain. J Biol Chem. 2006; 281(44): 33577–33587. PubMed Abstract | Publisher Full Text\n\nVerbeek HH, Alves MM, de Groot JW, et al.: The effects of four different tyrosine kinase inhibitors on medullary and papillary thyroid cancer cells. J Clin Endocrinol Metab. 2011; 96(6): E991–5. PubMed Abstract | Publisher Full Text\n\nCarlomagno F, Salvatore D, Santoro M, et al.: Point mutation of the RET proto-oncogene in the TT human medullary thyroid carcinoma cell line. Biochem Biophys Res Commun. 1995; 207(3): 1022–1028. PubMed Abstract | Publisher Full Text\n\nMatsubara D, Kanai Y, Ishikawa S, et al.: Identification of CCDC6-RET fusion in the human lung adenocarcinoma cell line, LC-2/ad. J Thorac Oncol. 2012; 7(12): 1872–1876. PubMed Abstract | Publisher Full Text\n\nSuzuki M, Makinoshima H, Matsumoto S, et al.: Identification of a lung adenocarcinoma cell line with CCDC6-RET fusion gene and the effect of RET inhibitors in vitro and in vivo. Cancer Sci. 2013; 104(7): 896–903. PubMed Abstract | Publisher Full Text\n\nSakamoto H, Tsukaguchi T, Hiroshima S, et al.: CH5424802, a selective ALK inhibitor capable of blocking the resistant gatekeeper mutant. Cancer Cell. 2011; 19(5): 679–690. PubMed Abstract | Publisher Full Text\n\nWatson AJ, Hopkins GV, Hitchin S, et al.: Dataset 1 in: identification of selective inhibitors of RET and comparison with current clinical candidates through development and validation of a robust screening cascade. F1000Research. 2016. Data Source"
}
|
[
{
"id": "14778",
"date": "05 Jul 2016",
"name": "Anderson J. Ryan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a screening cascade for the identification of potent RET inhibitors with selectivity against VEGFR2 (KDR) that will be of general interest in the field. The background to the research is well described, and the methods sufficiently detailed to allow others to replicate the experimental procedures and the conclusions are balanced.\n\nPoints\nEGFR is wrongly defined as ‘endothelial growth factor’\n\n..toxicity …resulting from off-target kinases. Should be on-target inhibition of non-RET kinases?\n\n…non-pharmacological toxicities…. Should be pharmacological toxicities associated with inhibition of non-RET targets?\n\nVEGF treatment does not affect levels of pRET\n\nHek293 is more commonly HEK293\n\n‘marvel’ is non-standard\n\nFigure 1. Error bars not defined, n=??. Should state this is RET enzyme assay\n\nFigures 2, 3, Suppl 2. Not stated what points represent [individual compounds??]\n\nTable 1. Confidence Interval values would be helpful as a surrogate of reproducibility, and support the claim of robustness, as reproducibility not specifically reported\n\nTable 1. Suggest including IC50 values for MZCRC1 cell proliferation above line for toxicity margin\n\nFigure 4. Not clear why some arrows in figure are there. Eg Arrow from in vivo PK to MZCRC1 POM/POP assay, or double-headed arrow from in vivo PK to in vitro DMPK\n\nSuppl Figure 1. No error bars (n=??)",
"responses": [
{
"c_id": "2088",
"date": "15 Jul 2016",
"name": "Mandy Watson",
"role": "Author Response",
"response": "Dear Anderson Thank you very much for reviewing this article. We appreciate your comments and suggested corrections and will address these in due course. Regards Mandy Watson"
}
]
},
{
"id": "14643",
"date": "07 Jul 2016",
"name": "Ian R. Hardcastle",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes the development and validation of the screening cascade for selective RET kinase inhibitors used to identify selective inhibitors described in Ref 1. The rationale for pursuing selective RET inhibitors is clearly explained, and the clinical limitations of current pan-kinase inhibitors that inhibit RET laid out.\nThe methods section has a well described rationale and is sufficiently detailed. The analysis of the data is sound and appropriate conclusions drawn. In particular, the method has accounted for the kinetics of binding, through time course experiments, and the possible impact of RET mutations in cell line studies.\n\nThe cascade developed adequately demonstrates selective cellular RET inhibition (POM). A good correlation is demonstrated between RET potency in vitro and growth inhibition in cell line models and this is taken as evidence of proof of mechanism.",
"responses": [
{
"c_id": "2087",
"date": "15 Jul 2016",
"name": "Mandy Watson",
"role": "Author Response",
"response": "Dear Ian Thank you very much for reviewing this article. Regards Mandy Watson"
}
]
},
{
"id": "14492",
"date": "11 Jul 2016",
"name": "Kevin Hudson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI have read this article and I support indexation. Given my past experience in kinase drug discovery I believe I have an appropriate level of expertise to comment on the article. The article is scientifically sound and is written with sufficient detail and clarity to allow others to reproduce the work.\nThe introduction section lays out a strong rationale for the efforts of this group to discover more selective RET inhibitors, indicating both the new emerging Oncology disease sector opportunities and the limited kinase selectivity of inhibitors currently in clinical use.\nThe results and discussion section clearly describes the screening cascade used to discover more selective RET inhibitors. The correlations illustrated between the different assays helps build confidence in the screening cascade. In addition, the screening cascade features different mutant forms (miss-sense, gene fusions) of RET in addition to WT, thereby indicating that compounds the authors discover from this cascade have potential application across a spectrum of tumours where RET is an oncogenic driver. The data derived from testing of clinical agents in their cascade reinforces both the rationale behind the work and helps validate the cascade.\n\nThe direction the cascade provides for producing more selective RET inhibitors is largely focused around minimizing activity against the KDR receptor tyrosine kinase. This makes sense given the chemical start point for this work was vandetanib, a proven KDR inhibitor. However, as chemical series evolve during a drug discovery project, there is the risk of introducing additional secondary pharmacology (inhibition of additional kinases beyond RET and KDR). The author’s use of a ‘Toxicity Margin’ calculation based around cell proliferation sensitivity of RET-null HEK293 cells provides a rapid and simple measure to track introduction of additional new pharmacology, although testing of frontrunner compounds in one of the widely available kinase panel screens may have been considered to fully appreciate the kinase selectivity of the project compounds.\n\nMinor comments / corrections:\nEGFR is an abbreviation for Epidermal Growth Factor Receptor, not endothelial growth factor as stated.\n\nVEGF treatment, not VEGFR treatment as stated\n\nThe text relating to Figure 1B could more clearly indicate this data came from pre- incubation and wash out experiments.\n\nFigure 2C may benefit from more clearly labeling them as enzyme assays, i.e. RET enzyme (M918T) vs RET enzyme (WT).\n\nOther pertinent comments proposing clarification of statistical measurements used in the figures, and in wording in paragraph 3 of the Introduction, have already been noted by other reviewer, Anderson J Ryan.",
"responses": [
{
"c_id": "2086",
"date": "15 Jul 2016",
"name": "Mandy Watson",
"role": "Author Response",
"response": "Dear Kevin, Thank you very much for reviewing this article. We appreciate your comments and suggested corrections and will address these in due course. Regards Mandy Watson"
}
]
},
{
"id": "14490",
"date": "11 Jul 2016",
"name": "Patrick A. Eyers",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis well-controlled study from Watson and colleagues manages to avoid many of the cliches used in reporting kinase inhibitor studies ('potent and selective', 'highly specific'), and will be useful for those in the field interested in on and off-targets of clinical, probe and intermediate compounds, especially with relevance to RET, which is fast becoming an interesting drug target.\nThe introduction is very clear, up-to-date with clinical and preclinical information, and leads into the new study efficiently.\nMinor points:\nOne sentence mentioning or placing in context the M918T (P+1 loop) and V804M (gatekeeper) cancer-associated RET mutants would be useful for readers, since one is employed here.\n\nIs Km[ATP] the same for M918T RET, given that it autophosphorylates much more efficiently in vitro (Plaza-Menacho et al., Mol Cell 53:738?). Is Km literature, or measured in-house?\n\nEGFR = Epithelial Growth Factor Receptor, not endothelial.\n\nWhat is the evidence that RET inhibition by bandetanib and cabozantinib is actually 'secondary pharmacology' in terms of in vivo efficacy? This statement might be tempered.\n\nCell POM: The POM assumes that the conversion of RET to pRET and KDR to pKDR in cells occurs inter/intramolecularly autophosphorylation (ie by the same protein kinase). Is this known, or surmised (reference would be useful).\n\nPlease add a citation for the MZ-CRC-1 cell line\n\nPlease add a citation, or Figure, demonstrating that HEK-293 cells do not express RET, since this is important for its use as control. What else is different between HEK-293 and MZ-CRC-1 though? Is this known?\n\nTable 1\nChallenging to distinguish between red and orange, how about another colour?\nWhich cell line is this data from?\n\nSupplementary Figure 2.\nAlthough ganetespib (HSP90i) destabilising effects provides a clear positive control, cabozantinib or sorafentib (sub micromolar RET inhibitors) also destabilise RET (and abolish autophosphorylation). But is this also true for the most potent in vitro inhibitors Ponatinib, Lenvatinib, PD16860/16964, since definition of the IC50 value must require measurement of the measured RET total vs. pRET ratio? Elimination of RET, in addition to inhibition of activity might be a nice dual-approach here, and might influence selection of candidates?",
"responses": [
{
"c_id": "2089",
"date": "15 Jul 2016",
"name": "Mandy Watson",
"role": "Author Response",
"response": "Dear Patrick Thank you very much for reviewing this article. We appreciate your comments and suggested corrections and will address these in due course. Regards Mandy Watson"
}
]
}
] | 1
|
https://f1000research.com/articles/5-1005
|
https://f1000research.com/articles/5-2043/v1
|
22 Aug 16
|
{
"type": "Research Note",
"title": "A reanalysis of “Two types of asynchronous activity in networks of excitatory and inhibitory spiking neurons”",
"authors": [
"Rainer Engelken",
"Farzad Farkhooi",
"David Hansel",
"Carl van Vreeswijk",
"Fred Wolf",
"Rainer Engelken",
"Farzad Farkhooi",
"Carl van Vreeswijk",
"Fred Wolf"
],
"abstract": "Neuronal activity in the central nervous system varies strongly in time and across neuronal populations. It is a longstanding proposal that such fluctuations generically arise from chaotic network dynamics. Various theoretical studies predict that the rich dynamics of rate models operating in the chaotic regime can subserve circuit computation and learning. Neurons in the brain, however, communicate via spikes and it is a theoretical challenge to obtain similar rate fluctuations in networks of spiking neuron models.\nA recent study investigated spiking balanced networks of leaky integrate and fire (LIF) neurons and compared their dynamics to a matched rate network with identical topology, where single unit input-output functions were chosen from isolated LIF neurons receiving Gaussian white noise input. A mathematical analogy between the chaotic instability in networks of rate units and the spiking network dynamics was proposed.\nHere we revisit the behavior of the spiking LIF networks and these matched rate networks. We find expected hallmarks of a chaotic instability in the rate network: For supercritical coupling strength near the transition point, the autocorrelation time diverges. For subcritical coupling strengths, we observe critical slowing down in response to small external perturbations. In the spiking network, we found in contrast that the timescale of the autocorrelations is insensitive to the coupling strength and that rate deviations resulting from small input perturbations rapidly decay. The decay speed even accelerates for increasing coupling strength.\nIn conclusion, our reanalysis demonstrates fundamental differences between the behavior of pulse-coupled spiking LIF networks and rate networks with matched topology and input-output function. In particular there is no indication of a corresponding chaotic instability in the spiking network.",
"keywords": [
"rate chaos",
"balanced state",
"mean field theory",
"network dynamics"
],
"content": "Introduction\n\nSlow neural dynamics are believed to be important for behavior, learning and memory (Churchland & Shenoy, 2007; Fee & Goldberg, 2011; Murray et al., 2014). Rate models operating in the chaotic regime show rich dynamics at the scale of hundreds of milliseconds and provide remarkable learning capabilities (Barak et al., 2013; Sussillo & Abbott, 2009; Toyoizumi & Abbott, 2011). Understanding the conditions of such a transition to chaos in more detailed network models has recently attracted a lot of interest (Harish & Hansel, 2015; Kadmon & Sompolinsky, 2015). However, neurons in the brain communicate via spikes and it is a challenge in computational neuroscience to obtain similar slow rate dynamics in networks of spiking neuron models.\n\nThis question was recently addressed in a paper by Ostojic (2014) published in Nature Neuroscience (Ostojic, 2014). It argues that an “unstructured, sparsely connected network of model spiking neurons can display two fundamentally different types of asynchronous activity”. When the synaptic strength is increased, networks of leaky integrate-and-fire (LIF) neurons would undergo a transition from the “well-studied asynchronous state, in which individual neurons fire irregularly at constant rates” to another “heterogeneous asynchronous state” in which “the firing rates of individual neurons fluctuate strongly in time and across neurons” (Ostojic, 2014). These two regimes would differ in an essential manner, the rate dynamics being chaotic beyond the phase transition. Finding a transition to chaotic slow-varying rate dynamics in spiking networks in such a simple model would be an important step towards an understanding of the computations underlying behavior and learning and would fill a gap in the current understanding of network dynamics. Here we re-examine the behavior of random LIF networks and demonstrate that there is no such phase transition to chaos in the spiking network analyzed in (Ostojic, 2014). While we confirm the observed deviation from the mean field theory description that assumes uncorrelated Gaussian fluctuations in time and among neurons, we controvert the validity of the presented analysis. We provide a series of tests of dynamical behavior that refute the existence of a chaotic instability and show that the analogy between the spiking network and the rate network is conceptually misleading and mathematically flawed.\n\nThe paper (Ostojic, 2014) starts with simulations of a network of LIF neurons for different values of the synaptic strength, J, while all other parameters are fixed to specific values. It is observed that the population mean firing rate of the neurons, ν0, is well described by a mean field calculation only below a certain coupling strength J*. At this value, the average firing rate starts to deviate from the mean field prediction more than 5%. (Figure 1a in (Ostojic, 2014), denoted Figure P1a; hereafter figures in (Ostojic, 2014) are denoted by their numbers preceded by a “P”). In (Ostojic, 2014), it was claimed that the “classical” asynchronous state exhibits an instability at J=J*. Above J* the dynamics would still be asynchronous, but in a way which would be essentially different from the “classical” asynchronous state. To assess this claim, the author replaced the full dynamics of the spiking LIF network by a rate model of similar connectivity, the “Poisson network”. Simulations indicate that as J increases, there is a value, J=Jc, at which the dynamics of the latter undergo a phase transition between a state in which the rates are constant in time (fixed point) and a state in which they fluctuate chaotically with long network generated time-scales. The author then derives an equation for a critical value Jc which is in agreement with the simulations of the Poisson model. For the parameters used in Figure P1 and P2 the value of Jc is rather close to J*. Apparently the author felt that this similarity, gives sufficient reason to justify two conclusions: (i) in the LIF network an instability occurs near J* which is of the same nature as the one occurring at Jc in the Poisson network. (ii) The asynchronous states below and above J* are essentially different in the LIF network.\n\n(a) Population averaged firing rate in the network vs. coupling strength J. Solid lines: Ricciardi mean field for C=1000 (red) and C=4000 (blue). Predictions for Jc (Equation 16) are indicated by the corresponding dashed vertical lines. Simulation results (event-based simulation implemented in Julia programing language) are also plotted. Dots: Δ=0.55 ms synaptic delay. Triangles: Δ=0.0 ms. Results for C=1000, N=10000 (red marker) and C=4000 and N=40000 (blue marker). (b) Averaged normalized AC of neuronal rate functions for J=0.8 mV and C=1000 (red) and C=4000 (dashed blue) LIF networks. The rate functions were computed by filtering the spike trains of the neurons (1 ms time bin) with a Gaussian filter with 10 ms (the thinnest lines), 50 ms (moderated lines) and 100 ms (the thickest lines) standard deviation. (c) Autocorrelation function of the spike trains (no filtering) normalized to the second pick. Solid lines: LIF network. Dashed lines: Poisson network. The results are shown for J = 0.5 mV (dark green), J=0.6 mV (dark orange), J = 0.7 mV (magenta) and J=0.8 mV (dark red). For the LIF the AC is also shown for J=0.4 mV (solid black). To compute the ACs for the Poisson network we simulated a network for 100 s (time step 1 ms) and averaged the results over 40 realizations of the initial conditions. The network size is N=100000 for 0.5≤J≤0.6mV and N=10000 for J>0.6 mV. For the LIF network we averaged spike autocorrelation of 3000 randomly chosen neurons with a 1 ms bin following Equation 23 in the paper. All parameters are as in Figure P3. (d) Subcritical behavior of the systems. Rate network and spiking network are both perturbed in the constant feed-forward input current µ0 in the least stable direction of the linearized rate dynamics (Equation 16) for different coupling strengths J. The resulting rate deviation is projected onto the perturbation direction. Dashed lines reflect the normalized decay of this perturbation in the rate network and the solid lines those of the spiking network (averaged over 1.42 million perturbations). The perturbation was applied to the constant feed-forward input µ0 for 2 ms where the standard deviation of the perturbation vector was 1 mV. Longer perturbation durations (10 ms) and weaker perturbation strengths (standard deviation 0.1 mV) gave very similar results (not shown). Perturbation direction, strength, duration and network realization were exactly the same for rate and spiking network. Other parameters as in (Ostojic, 2014).\n\nHowever, as we now show, the reported agreement between the predicted transition at Jc and the spiking network simulation results is coincidental and only valid for the chosen parameters used in the paper (Ostojic, 2014) but not in general. We start by providing two counter-examples to statements (i) and (ii).\n\n\nMethods and results\n\nOur first counter-example is the LIF model considered in (Ostojic, 2014), we take N=40000 neurons and C=4000 synapses per neuron instead of N=10000 and C=1000 (all other parameters as in Figure P1, except for the network size, keeping the connection probability constant). The population firing rate, ν0 (J), is plotted in Figure 1a. It deviates from the mean field prediction at J*≅0.3 mV by more than 5%. Nonetheless, the critical point in the corresponding Poisson rate network is Jc≅0.96 mV and thus it is more than three times larger than J*.\n\nOur second counter-example is the LIF network of Figure P1 and P2 with the same parameters except for the delay, Δ. We note that the delay does not affect the existence of the asynchronous state and importantly plays no role in the mathematical considerations of Ostojic (2014). As these yield identical results irrespective of delay we consider the simplest case: Δ = 0 ms. Strikingly, the spiking network shows no longer a large deviation from the mean-field prediction (Figure 1a). However, the proposed analogy with the Poisson rate network still predicts that a deviation should occur at J*≅0.49 mV, since the transition to chaos in the Poisson network is independent of the delay. The author seems to be somewhat aware of this discrepancy. Indeed, it is stated in the Online Methods that delays must be larger than the refractory period, because “if the delays are shorter, spikes that reach a neuron while it is refractory do not have an effect and the overall coupling is effectively reduced” (Ostojic, 2014). If this was correct, this effective reduction should be reflected in the formula for predicting Jc (Equation 16). This is not the case: the latter does not depend on Δ. In addition, the spiking network for Δ = 0 ms in fact exhibits no increased level of network synchrony measured by the common synchrony measure X (Figure 2a) (Hansel & Mato, 2003).\n\n(a) Synchrony measure X vs. coupling strength J. Dots: Δ=0.55 ms. Triangles: Δ=0.0 ms for N=10000, C=1000. X is defined as in (Hansel & Mato, 2003) on the phases of neurons. Note that zero delay does not increase network synchrony. (b) Coefficient of variation of the interspike intervals vs. coupling strength J. (c) Distribution of membrane potentials for different coupling strength J (in mV). (d) Example voltage trace for J=0.8 mV shows very negative excursions followed by short bursts of action potentials. Red dots indicate spike times. Numerically exact event-based simulation were implemented in Julia programing language. Other parameters are chosen as in (Ostojic, 2014).\n\nIt is also argued in the paper (Ostojic, 2014) that the results plotted in Figure P3a and b support the analogy between the rate dynamics of the Poisson model and the dynamics of the LIF network. However, the comparison made in this figure is conceptually misleading. In the Poisson model, the rate as a function of time is an unequivocally defined quantity. It is the dynamical variable of the model and the time scale over which the rate fluctuates for strong enough coupling is fully determined by these dynamics. This is not the case in the LIF model where the “rate” and its “dynamics” depend on the temporal width over which the spiking activity is filtered. The width of the Gaussian filter used in (Ostojic, 2014) is 50 ms. This choice is arbitrary and is the reason for the similarity observed in the rate autocorrelations (ACs) plotted in the upper and lower panels in Figure P3b which depends on this choice (Figure 1b). The rate functions were computed by filtering the spike trains of the neurons (1 ms time bin) with a Gaussian filter with 10 ms (the thinnest lines), 50 ms (moderated lines) and 100 ms (the thickest lines) standard deviation. Moreover, the spike ACs plotted in Figure P3c for the two models exhibit essential differences as we now show.\n\nFor J=0.2 and 0.4 mV, the spike AC in the Poisson rate model (Figure P3c, upper panel) is close to a Dirac function reflecting that the dynamics are at fixed point - that is the rate variable from which the Poisson process of the spikes is generated is constant. For J=0.6 mV the spike AC is very different: a broad component has now appeared. It is flat at zero time lag and has a negative curvature at short time lags (Figure 1c and Figure P3c). A detailed analysis reveals that this change has all the characteristics of a true phase transition. It shows that close to the phase transition, the amplitude vanishes proportionally to J-Jc and the decorrelation time diverges as 1/J–Jc (Figure S1a). To compute the ACs for the Poisson network we simulated a network for 100 s (time step 1 ms) and averaged the results over 40 realizations of the initial conditions. The network size is N=100000 for 0.5≤J≤0.6mV and N=10000 for J>0.6 mV. For the LIF network we averaged spike autocorrelation of 3000 randomly chosen neurons with a 1 ms bin following Equation 23 in the paper. All parameters are as in Figure P3.\n\nThe spike AC behaves very differently in the LIF network. For J=0.2 mV it exhibits at zero time lag a sharp peak flanked by a trough which reflects the refractoriness (absolute and relative) of the single neuron dynamics. As J increases, there is a progressive change in the AC shape. Eventually, the trough disappears. The flanks of the zero peak are now decreasing exponentially (Figure 1c, solid lines). A careful analysis reveals that the typical time constant of this decrease depends only weakly on J (Figure 1c, solid lines). It is always on the order of the membrane time constant of the neurons (20 ms). Note also that by contrast with what is observed in the Poisson network, for J=0.5 to 0.8 mV, the spike AC curvature is always positive and peaked around zero time lag (Figure 1c, dashed lines).\n\nHow do the “strong fluctuations” in the “heterogeneous regime” emerge? For increasing J, the spiking activity of single neurons becomes increasingly irregular, quantified by the mean coefficient of variation (cv) of the interspike interval distribution (Figure 2b). At the same time, the distribution of membrane potentials develops a very long tail towards negative voltages (Figure 2c). For strong coupling (J=0.8 mV), voltage traces of individual neurons show long very negative voltage excursions, followed by short bursts of action potentials (Figure 2d). This explains the super-Poissonian irregularity (CV>1). The super-Poissonian nature of spiking irregularity and the unphysiological negative voltage deviations are properties related to the linear V̇-V-relationship of the LIF model. A mean-field description of this phenomenon requires self-consistent spike train autocorrelations (Lerchner et al., 2006; Wieland et al., 2015). For other integrate-and-fire neurons e.g. the quadratic-integrate-and-fire model, even for very strong coupling J, e.g. J = 20 mV, the mean coefficient of variation does not increase beyond one and no strongly negative voltage excursions are observed. All parameters are as in Figure P1.\n\nAdditionally, in order to compare the behavior of spiking and rate models below the postulated phase transition, we perturbed rate and spiking networks of identical topology in the least stable direction of the linearized rate dynamics, predicted by Equation 16 in the paper (Ostojic, 2014). The resulting rate deviation is projected onto the perturbation direction. The perturbation was applied to the constant feed-forward input µ0 for 2 ms where the standard deviation of the perturbation vector was 1 mV. Figure 1d shows that the decay of the perturbation in the rate network slows down near the transition, indicating a critical slowing down (Figure 1d, dashed lines). If there were a “mathematically analogous” transition in the spiking network, also its perturbation should decay slower as the transition is approached. Our result (Figure 1d, solid lines) shows that the decay time-scales of the perturbation (averaged over 1.42 million perturbations) is insensitive to J and it stays close to the membrane time constant (similar to solid lines in Figure 1c). Longer perturbation durations (10 ms) and weaker perturbation strengths (standard deviation 0.1 mV) gave very similar results (not shown). All other parameters are chosen as in (Ostojic, 2014).\n\n\nConclusion\n\nWe therefore conclude that, contrary to what was argued by the author, the spiking LIF network studied in (Ostojic, 2014) does not exhibit a phase transition to a chaotic state similar to the one occurring in the studied rate model. The reported mismatch between the average firing rate in this LIF network simulations and the mean-field calculation is unrelated to such a transition.\n\n\nData and software availability\n\nZenodo: Reanalysis of “Two types of asynchronous activity in networks of excitatory and inhibitory spiking neurons”, 10.5281/zenodo.59624 (Engelken & Farkhooi, 2016).\n\nThe data for this article are also available on the Open Science Framework at: https://osf.io/q3vt4/",
"appendix": "Author contributions\n\n\n\nAll authors participated in the research design. RE wrote the simulation and analysis code for Figure 1a,d and Figure 2a–d. FF wrote the simulation and analysis code for Figure 1b,c and Figure S1a,b. Earlier simulations were performed by RE, FF, DH. All authors participated in the interpretation of the results and in manuscript writing. DH, CvW and FW share joint first authorship of this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe work of DH and CvV was partially supported by grants ANR-13-BSV4-0014-03-BALAV1 and ANR-14-NEUC-0001-01-BASCO and performed in the framework of the France-Israel Laboratory of Neuroscience (FILN). FF was supported by the BMBF, FKZ 01GQ 1001B. RE and FW received funding from Evangelisches Studienwerk Villigst, DFG through CRC 889 and Volkswagen Foundation.\n\n\nAcknowledgments\n\nWe thank Ran Darshan, Omri Harish and Gianluigi Mongillo for fruitful discussions.\n\n\nSupplement material\n\n(a) The decorrelation time (τ0, violet diamond, left y-axis) and amplitude at zero time lag (beta, orange circles, right y-axis) of the baseline-subtracted population averaged spike AC are plotted vs. J for the Poisson network. These parameters were obtained by fitting the spike AC with ACF(τ) = β/cosh(τ/τ0)2 (see Figure S1b). Inset: the rescaled estimated τ0-2 (left axis, violet) and β values (orange, right axis) for J=0.5, 0.5125, 0.525, 0.5375, 0.55, 0.5625, 0.575, 0.5875 and 0.6 mV, to show that they vanish linearly near the phase transition. (b) The non-normalized spike AC can be very well fitted by ACF(τ) = β/cosh(τ/τ0)2. Dashed lines: Simulation results; J=0.525 mV (cyan, right y-axis) and J=0.8 mV (dark red, left y-axis): Black solid line: The fits.\n\n\nReferences\n\nBarak O, Sussillo D, Romo R, et al.: From fixed points to chaos: three models of delayed discrimination. Prog Neurobiol. 2013; 103: 214–222. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChurchland MM, Shenoy KV: Temporal complexity and heterogeneity of single-neuron activity in premotor and motor cortex. J Neurophysiol. 2007; 97(6): 4235–4257. PubMed Abstract | Publisher Full Text\n\nEngelken R, Farkhooi F: reanalysis of “Two types of asynchronous activity in networks of excitatory and inhibitory spiking neurons”. Zenodo. 2016. Data Source\n\nFee MS, Goldberg JH: A hypothesis for basal ganglia-dependent reinforcement learning in the songbird. Neuroscience. 2011; 198: 152–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHansel D, Mato G: Asynchronous states and the emergence of synchrony in large networks of interacting excitatory and inhibitory neurons. Neural Comput. 2003; 15(1): 1–56. PubMed Abstract | Publisher Full Text\n\nHarish O, Hansel D: Asynchronous Rate Chaos in Spiking Neuronal Circuits. PLoS Comput Biol. 2015; 11(7): e1004266. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKadmon J, Sompolinsky H: Transition to chaos in random neuronal networks. Phys Rev X. 2015; 5(4): 041030. Publisher Full Text\n\nLerchner A, Ursta C, Hertz J, et al.: Response variability in balanced cortical networks. Neural Comput. 2006; 18(3): 634–659. PubMed Abstract | Publisher Full Text\n\nMurray JD, Bernacchia A, Freedman DJ, et al.: A hierarchy of intrinsic timescales across primate cortex. Nat Neurosci. 2014; 17(12): 1661–1663. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOstojic S: Two types of asynchronous activity in networks of excitatory and inhibitory spiking neurons. Nat Neurosci. 2014; 17(4): 594–600. PubMed Abstract | Publisher Full Text\n\nSussillo D, Abbott LF: Generating coherent patterns of activity from chaotic neural networks. Neuron. 2009; 63(4): 544–557. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToyoizumi T, Abbott LF: Beyond the edge of chaos: amplification and temporal integration by recurrent networks in the chaotic regime. Phys Rev E Stat Nonlin Soft Matter Phys. 2011; 84(5 Pt 1): 051908. PubMed Abstract | Publisher Full Text\n\nWieland S, Bernardi D, Schwalger T, et al.: Slow fluctuations in recurrent networks of spiking neurons. Phys Rev E Stat Nonlin Soft Matter Phys. 2015; 92(4): 040901. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15814",
"date": "07 Sep 2016",
"name": "Jonathan Touboul",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nCharacterizing the dynamics of spiking neural networks and their transitions is currently a prominent issue in computational neuroscience. The question is largely non-trivial and riddled with subtleties: numerical simulations are often delicate to interpret, and theoretical tools to analyze these dynamics are still being developed.\n\nThe theoretical community have devoted important effort to address this question. Indeed, progresses on this question would advance our understanding of the brain and its computations. A question of particular interest is to characterize phase transitions to chaotic regimes in spiking networks. Indeed, since the seminal work of Sompolinsky, Crisanti and Sommers1 on rate networks, chaotic regime were shown to have rich dynamics able to support efficient computations and learning (see e.g. Sussillo & Abbott 20092). Whether spiking networks do show a similar transition and thus share similar properties as rate networks has recently been the focus of several researches and is an important endeavor in computational neuroscience3.\nThe present paper addresses a few important questions on the interpretations and conceptual approach of a theoretical article appeared in Nature Neuroscience in 20144 dealing precisely with dynamics and transitions in spiking networks. That paper argued for the existence a transition in a balanced spiking network, between an asynchronous and a \"new highly fluctuating regime\", using in particular transitions of an associated rate model. The present article comes back to this comparison between spiking and rate network, and argues that, in contrast with the approach of Ostojic (2014)4, it is not possible to extract accurate information on the spiking network from an analysis of the particular rate network studied, by showing a mismatch between their qualitative dynamics. Moreover, using relevant numerical quantities to identify phase transitions (synchrony measure and effect of perturbations), the authors establish the absence of phase transition in the spiking network, while the rate network does present the hallmarks of phase transitions.\n\nThe present paper thus contributes to an important scientific debate on the characterization of the dynamical regimes of spiking networks. This question has been the topic of very recent important works that advance our understanding of spiking networks and the associated mean-field limits (to cite a few, see Kadmon & Sompolinsky 20155, Harrish & Hansel 20156, Goedeke, Schuecker & Helias 20167).\n\nFor its contribution to the scientific debate on a timely and important topic in theoretical neuroscience, this paper shall be helpful to the readers interested in the existence and nature of transitions in spiking networks.",
"responses": []
},
{
"id": "16565",
"date": "27 Sep 2016",
"name": "Maoz Shamir",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe origin and possible computational role of neuronal noise has been the focus of considerable scientific effort during the past decades. In particular, the transition to chaos has been extensively studied using simplified rate-models and much is known about this transition. A recent work studied the dynamics of a sparsely connected network of excitatory and inhibitory spiking neurons in the balance regime and a compelling mapping between spiking neural network and rate model was proposed. Following analysis and numerical simulations it was suggested that a novel type of asynchronous state exists and it was further hypothesized that this novel state is useful for complex information processing in the central nervous system. Here the authors reevaluate this claim. Several counter examples are provided to prove that the claim of a transition to a second type of asynchronous state does not hold. Furthermore, the origin of ‘strong fluctuation’ super-Poisson irregular firing is studied and is found to be related to extremely negative voltage fluctuations that are beyond the typical physiological range. The work is timely. The results are solid and well presented. I also find the effort devoted to reproduce and re-evaluate results refreshing. I believe this paper will contribute to the scientific debate.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2043
|
https://f1000research.com/articles/5-1435/v1
|
20 Jun 16
|
{
"type": "Opinion Article",
"title": "Inhibitory system overstimulation plays a role in the pathogenesis of neuromuscular and neurological diseases: a novel hypothesis",
"authors": [
"Bert Tuk"
],
"abstract": "Based upon a thorough review of published clinical observations regarding the inhibitory system, I hypothesize that this system may play a key role in the pathogenesis of a variety of neuromuscular and neurological diseases. Specifically, excitatory overstimulation, which is commonly reported in neuromuscular and neurological diseases, may be a homeostatic response to inhibitory overstimulation. Involvement of the inhibitory system in disease pathogenesis is highly relevant, given that most approaches currently being developed for treating neuromuscular and neurological diseases focus on reducing excitatory activity rather than reducing inhibitory activity.",
"keywords": [
"Neuromuscular disease",
"neurodegeneration",
"ALS",
"FTD",
"Alzheimer’s disease",
"Parkinson’s disease",
"Huntington’s disease",
"Primary Lateral Sclerosis"
],
"content": "The clinical manifestations of neuromuscular and neurological diseases have high overlap\n\nThe pathogenesis of most neuromuscular and neurological diseases is poorly understood, despite their devastating impact on quality of life and the fact that they were first described more than a century ago. Clinically, neuromuscular diseases manifest as progressive muscle weakness together with a general set of motor symptoms, including speech-related difficulties, impaired mobility, and reduced fine motor skills1. In contrast, neurological diseases manifest primarily as a progressive decline in cognitive function. Interestingly, the clinical manifestations of neuromuscular and neurological diseases also overlap; this overlap is summarized in Table 1 for primary lateral sclerosis (PLS), amyotrophic lateral sclerosis (ALS), ALS with frontotemporal dementia (ALS-FTD), FTD with ALS (FTD-ALS), FTD, Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease1–29. The clinical features shared between the neuromuscular disease ALS and the neurological disease FTD exemplify this overlap, as late-stage ALS can lead to the manifestation of FTD; conversely, FTD can progress to ALS, leading to the manifestation of FTD-ALS2–8.\n\n√, present; –, absent in most patients\n\nPLS, primary lateral sclerosis; ALS, amyotrophic lateral sclerosis; FTD, frontotemporal dementia; CSF, cerebrospinal fluid\n\n\nElevated glutamate levels are involved in the pathogenesis of both neuromuscular and neurological diseases\n\nA key observation gleaned from analyzing Table 1 is the finding that glutamate levels are increased in the cerebrospinal fluid (CSF) of patients in all eight diseases22–29. Glutamatergic (i.e., excitatory) overstimulation induces excitotoxicity in cultured neurons and is believed to be an important factor in the pathogenesis of both neuromuscular and neurological diseases22–29. Glutamate-induced excitotoxicity can result in the decay of neuronal pathways that innervate muscles and other physiological systems22–29. This decay gives rise to the loss of physiological function and is considered to lead to the clinical manifestations that present with both neuromuscular disease and neurological disease22–29. I hypothesize that these increased glutamate levels are actually a homeostatic response to an overstimulated inhibitory system. This novel hypothesis is based upon the observation that the clinical findings in neuromuscular and neurological diseases can be explained by inhibitory activity, as discussed below.\n\n\nDespite increased glutamate levels, patients with neuromuscular and neurological diseases do not have increased epileptic activity\n\nSince they were first diagnosed more than a century ago, the clinical manifestations of neuromuscular and neurologic diseases have been well described. Strikingly, however, the consequences of one key clinical feature of these diseases—the absence of an elevated risk of seizure activity—have been largely overlooked.\n\nThis is exemplified for ALS in which a broad, detailed retrospective study of the medical records of 657 ALS patients revealed that none of the patients presented with epilepsy as a co-morbid condition9. Moreover, a thorough search of PubMed for articles published from 1966 through 2016 using the key words “seizure” or “epilepsy” in combination with “amyotrophic lateral sclerosis” or “ALS” confirms the striking absence of epilepsy and/or seizures in ALS patients. This finding is consistent with the absence of seizures and/or epilepsy in review articles describing the clinical manifestation of ALS2–6.\n\nA key observation that makes the absence of seizure activity in ALS even more remarkable is increased glutamate levels in the cerebrospinal fluid (CSF) of patients with ALS; on average, glutamate levels in the CSF of ALS patients are increased by 100%, and some ALS patients can have an increase of up to 800%23. Importantly, increased glutamate levels are generally associated with epileptic seizures30,31. Thus, given the increased glutamate levels typically measured in the CSF of ALS patients, one would logically expect that the prevalence of epilepsy in ALS patients should be elevated relative to the general population. However, despite this expectation, epileptic seizures are simply not reported among ALS patients.\n\nStrikingly, in addition to ALS, none of the other seven diseases listed in Table 1 typically present with an increased risk of epileptic seizures, either2–15, even though all eight diseases present with elevated glutamate levels in the CSF22–29.\n\n\nDespite elevated glutamate levels, muscles in neuromuscular and neurological patients are inhibited\n\nA second key observation is that neuromuscular and neurological diseases have an inhibitory effect on muscle function, rather than being excitatory. The diseases listed in Table 1 are characterized by muscle inhibition, even though glutamate—which, as discussed above, is generally increased in these diseases—is the major neurotransmitter that drives muscle activation by increasing the firing rate of motor neurons. Remarkably, however, despite having increased levels of glutamate in the CSF, patients with neuromuscular and neurological diseases do not have increased muscle activation. This is exemplified most clearly by ALS, a disease with highly elevated glutamate levels22,23 and complete muscle inhibition in the end stages. Although fasciculation and/or cramps can be observed in ALS patients2–4, these features occur in debilitated muscles as they progress from a fully functional state toward a fully inhibited state.\n\nOne possible explanation for these seemingly contradictory findings is a second system that exerts strong anticonvulsive activity in both neuromuscular and neurological diseases. Importantly, such a system should be as widespread throughout the nervous system as the glutamatergic system, and the inhibitory system fulfills these requirements. Specifically, the inhibitory (GABA) system i) functions to oppose the glutamatergic excitatory neurotransmitter system, ii) inhibits muscle activity by reducing the firing rate of motor neurons, and iii) exerts strong anticonvulsive activity26,30,31.\n\n\nThe clinical features of neuromuscular and neurological diseases can be induced by increasing inhibitory activity\n\nA third key observation is that the clinical manifestations of neuromuscular and neurological diseases can be induced using interventions that increase GABAergic (i.e., inhibitory) activity (Table 2). For example, activating the GABAergic inhibitory system using benzodiazepines can render healthy muscles dysfunctional32,33. In addition, fatal respiratory depression can be induced by administering an overdose of the GABAergic benzodiazepine midazolam34. Chronically stimulating the inhibitory system can cause chronic muscle disuse that can lead to muscle atrophy35. Moreover, ingestion of alcohol (another GABAergic inhibitory compound36) impedes coordination and causes slurred speech (dysarthria), which are features of neuromuscular and neurological diseases. In cats, dysphagia (difficulty swallowing) can be either induced or reversed using GABA agonists or GABA antagonists, respectively37. Dysphagia has also been reported in humans following the administration of either benzodiazepines38–40 or alcohol41. Administration of benzodiazepines reduces voluntary saccadic eye movement function42 and increases EEG beta-wave activity42, clinical manifestations that also occur in neuromuscular and neurological diseases18,43,44. Increased GABAergic inhibitory activity can also cause bladder45,46 and gastrointestinal dysfunction47,48, both of which can manifest in neuromuscular and neurological diseases19–21. Strikingly, GABAergic activity can also explain the overlapping clinical manifestations between Alzheimer’s disease and alcohol-related dementia49, and it can explain the increase in dementia-like symptoms observed after the administration of the benzodiazepine diazepam50. Inhibitory activity can also explain neuromuscular and neurological disease predisposition in the elderly, as the sensitivity to GABA inhibitory activity is known to increase with age51. Finally, GABAergic activity has been implicated in cognitive dysfunction52–54, which is a hallmark feature of neurological diseases and is often observed in late-stage neuromuscular disease2–7. Taken together, these findings support the notion that the clinical features associated with neuromuscular and neurological diseases can be induced by activating the inhibitory system.\n\n\nModulating inhibitory activity can explain the progression of ALS in clinical trials\n\nIn addition to mimicking the majority of clinical manifestations observed in neuromuscular and neurological diseases, GABAergic activity can also explain the more rapid disease progression of ALS reported in clinical trials in which patients received GABAergic compounds. For example, in two trials gabapentin increased the rate of disease progression in patients with ALS55. A similar effect was reported in patients with ALS who received the GABAergic compound topiramate56. GABAergic action can also explain the more rapid disease progression of ALS in clinical trials in which patients received the antibiotic minocycline57, which has GABAergic activity58. Finally, GABAergic involvement can explain the observed efficacy of the taurine conjugate form of ursodeoxycholic acid (UDCA) in ALS patients59, as UDCA inhibits GABAergic action60.\n\n\nNeuromuscular and neurological manifestations can be attributed to simple inhibition and/or recurrent inhibition\n\nA fourth key observation is that the clinical manifestations associated with neuromuscular and neurological diseases can be attributed to the activity of either simple inhibition (SI) or recurrent inhibition (RI) pathways. Specifically, I postulate that differences between muscles under the control of SI and/or RI underlie the important—yet poorly understood—manifestations of neuromuscular and neurological diseases.\n\nThe inhibitory system functions via both SI and RI61. The RI system controls physiological functions that play a role in counteracting gravitational forces and other external forces acting on the body. During locomotion and/or to counteract the effects of gravity, RI uses a negative inhibitory feedback loop (Figure 1), thereby providing muscles with additional, stabilizing input. Therefore, muscles involved in movement and lifting heavy objects are subject to RI. Examples of RI-innervated muscles include the limb and thorax muscles, as well as the neck muscles that control head movement.\n\nWith recurrent inhibition (RI), input from descending pathways (DP) reaches the motor neuron (MN). In response, the MN activates the target myocyte; in addition, the MN also activates Renshaw cells (RC), which then inhibit the motor neuron through a negative feedback loop.\n\nIn contrast, neuronal pathways that do not play a role in locomotion or counteracting gravity selectively utilize SI61. Examples of SI-innervated muscles include facial, speech, pharyngeal, and eye muscles, as well as muscles that are involved in bowel and bladder function.\n\nTable 3 summarizes the involvement of SI and RI in the principal clinical manifestations of neuromuscular and/or neurological diseases. A close examination of Table 3 reveals that one set of muscles—namely, the respiratory muscles—is innervated by both SI and RI pathways61,62. This dual innervation arises because the respiratory muscles play a role in both respiratory function and maintaining body posture61.\n\n\nSI and RI involvement can account for various onset manifestations in ALS\n\nStrikingly, the categorization between SI-innervated and RI-innervated muscles coincides with the categorization of muscles affected in limb-onset ALS, bulbar-onset ALS, and respiratory-onset ALS. Approximately 70%, 25%, and 5% of ALS patients present initially with limb involvement (limb-onset ALS), bulbar symptoms (bulbar-onset ALS), or respiratory symptoms (respiratory-onset ALS), respectively2, and this difference in onset can be explained by differences in SI versus RI involvement (Table 4). Patients with both SI and RI involvement at the onset of disease present with respiratory-onset ALS. In contrast, patients with primarily SI involvement present with bulbar-onset ALS, whereas patients with primarily RI involvement present with limb-onset ALS (see Table 4). Thus, SI and RI differentiation can account for this difference in ALS onset.\n\n\nSI versus RI involvement can account for differences in life expectancy among patients with ALS\n\nThe differential involvement of the SI or RI system can also explain the observed differences in life expectancy between patients who present with limb-onset, bulbar-onset, or respiratory-onset ALS. Specifically, patients with respiratory-onset ALS generally have the shortest life expectancy following diagnosis4. As discussed above, increased activity of both the SI and RI pathways leads to fatal respiratory depression, the principal cause of early death in patients with ALS.\n\nIncreased activity of either the SI or RI pathway—but not both—can also lead to respiratory depression, albeit not to fatal levels. Under these conditions, respiratory function, though impaired, can be maintained by either SI or RI pathway activity. However, due to impaired respiratory function resulting from either SI or RI overstimulation (see Table 4), these patients can die from dysphagia-related malnutrition and/or aspiration pneumonia. Thus, patients with either SI or RI involvement—but not both—generally live longer than patients with both SI and RI involvement. Importantly, this observation can also explain why patients with motor neuron diseases at either end of the SI/RI spectrum—for example, primary lateral sclerosis, progressive muscular atrophy, progressive bulbar palsy, or pseudobulbar palsy—have a longer life expectancy than patients with ALS4, which lies in the middle of the spectrum.\n\n\nThe SI and RI pathways can explain both the progression of ALS into FTD and the progression from FTD into ALS\n\nDifferences between the effects of SI versus RI involvement can explain the fact that although ALS and FTD generally involve two distinct systems, these two diseases have a certain degree of overlap with respect to their clinical manifestations (see Table 1). Thus, if RI overstimulation precedes SI involvement, the patient can present with an initial diagnosis of ALS and can progress to ALS-FTD, a common manifestation of cognitive dysfunction observed in 20–50% of patients with late-state ALS patients2–7. Alternatively, if SI overstimulation precedes RI involvement, FTD is the initial diagnosis, and the disease can progress to FTD-ALS when the RI pathway becomes involved. Moreover, the division between SI and RI can also explain the overlap between subcategories of ALS and FTD with respect to impaired cognition and altered behavior that involve SI, and movement dysfunction that involves RI.\n\n\nDifferential involvement of SI and RI can account for the wide variety of clinical manifestations in ALS\n\nAlthough it is generally considered one disease, ALS can present with a wide spectrum of clinical manifestations, and this spectrum can be explained by the involvement of SI and/or RI pathways. For example, SI overstimulation can lead to bulbar, cognitive, and frontotemporal dementia-related manifestations without causing severe muscle wasting or respiratory malfunction (for example, as observed in patients with bulbar-onset ALS). On the other hand, RI overstimulation can lead to locked-in syndrome, a state in which the patient retains cognitive and emotional function but becomes “locked” in their body, with all of the muscles that counteract gravity and other external forces rendered essentially dysfunctional. Interestingly, the only muscles that are spared in locked-in syndrome—and the only way in which end-stage patients can communicate with the outside world—are the muscles that control eye movement. This is an important observation, given that the muscles that control eye movement are not controlled by RI pathways (see Table 3). The distinction between SI and RI can also explain the observation that some patients with ALS have fully intact cognitive and emotional functions even after their muscles involved in countering gravity have become dysfunctional; the most famous example of this phenomenon is Stephen Hawking, who despite being diagnosed with ALS in his early twenties remains active as a prominent theoretical physicist, now in his seventies63.\n\n\nSplit-hand syndrome in ALS can be explained by differential innervation of SI and RI pathways\n\nSplit-hand syndrome is common among patients with ALS64. With split-hand syndrome, the abductor pollicis brevis (APB) and first dorsal interosseous (FDI) muscles are affected, whereas the abductor digiti minimi (ADM) muscle is relatively spared. This syndrome is particularly puzzling, as these muscles are innervated identically, yet are affected differently64. I propose that split-hand syndrome can be attributed to differences in the extent to which RI pathways innervate the hand muscles that are involved in precision gripping, versus muscles that also play a role in power gripping. With precision gripping (for example, when using a pen), the fingers and thumb press against each other; this type of grip does not involve lifting a relatively heavy object65. In contrast, power gripping (for example, when gripping a hammer or lifting a heavy pan) uses the fingers, palm, and thumb to clamp down on a heavy object in order to lift and control the object65. Napier65 used this distinction to distinguish muscle activities that are involved in body locomotion and/or posture from muscle activities that do not involve locomotion or posture. Thus, Napier’s separation also categorizes muscle activities into those that are controlled by RI and those that are controlled by SI. Because the primary function of the ADM muscle is to move the little finger (i.e., the fifth digit) away from the hand, the ADM muscle is only involved in precision gripping and would therefore not be affected by RI overstimulation. This is consistent with the reported absence of RI in motor neurons that innervate the ADM61,66. On the other hand, the APB and FDI muscles are involved in the opposition and extension of the thumb and are therefore involved in power gripping61; thus, these two muscles are affected by RI overstimulation.\n\n\nParkinson’s disease rest tremors can be attributed to differences between SI and RI involvement\n\nIn Parkinson’s disease, rest tremors arise from involuntary rhythmic oscillatory movements of a body part at rest; these tremors stop when the patient actively moves the affected body part. The pathways that underlie rest tremors have not been identified, and the fact that rest tremors resolve during voluntary movement is one of the most puzzling observations associated with Parkinson’s disease67. However, because these tremors occur at rest (and not during active motion or while countering the effects of gravity), the muscles involved are likely innervated by SI pathways, making rest tremors an SI-specific phenomenon. This is further illustrated by the finding that rest tremors resolve when the affected body part becomes involved in locomotion, stance, or maintaining inertia67, actions that involve muscles that are controlled by RI61. Interestingly, the hand tremor that is most specific to patients with Parkinson’s disease—the so-called “pill-rolling tremor”—also results from muscles that are innervated solely by SI pathways. Specifically, the pill-rolling tremor involves muscles that play a role in precision gripping but not in power gripping, a distinction that is highly reminiscent of split-hand syndrome in ALS (see above). Furthermore, involvement of the inhibitory system in Parkinson’s disease rest tremors is supported by the observation that the rest tremors observed in restless legs syndrome resolve after the administration of quinine (FDA Drug safety communication, 2010), a compound that reduces inhibitory activity68.\n\n\nThe differentiation of clinical manifestations in neuromuscular and neurological diseases can be attributed to SI versus RI pathways\n\nThe differentiation between SI and RI summarized in Table 3 can explain the three categories of fatal symptoms that arise in end-stage neuromuscular and neurological diseases (Table 5). One striking observation from Table 5 is that both SI and RI can be attributed to fatal respiratory failure, the major cause of death among ALS patients. Overstimulation of SI pathways leads to bowel dysfunction, bladder dysfunction, and dysphagia-related malnutrition and aspiration pneumonia; these symptoms are the major causes of death among patients with FTD, Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease. On the other hand, overstimulation of RI pathways can lead to end-stage locked-in syndrome.\n\nDifferential SI and RI involvement can also account for the wide variety of clinical manifestations in neuromuscular and neurological diseases during disease progression. As a group, neuromuscular and neurological diseases present with a wide spectrum of clinical manifestations (see Table 1), and stimulation of SI and/or RI pathways can account for this spectrum. For example, SI overstimulation can lead to FTD, Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease, whereas RI overstimulation can lead to locked-in syndrome. Finally, overstimulation of both the SI and RI pathways can lead to ALS.\n\nFinally, the differentiation between SI and RI can help explain the differences in life expectancy among patients with various neuromuscular and neurological diseases. As discussed above, increased activity of both the SI and RI pathways leads to fatal respiratory depression (see Table 5), the principal cause of death in patients with ALS, the neuromuscular disease with the shortest life expectancy. Increased activity of either the SI or RI pathway—but not both—can also lead to respiratory depression, albeit not to direct fatal levels. Thus, patients with either SI or RI overstimulation generally live longer than patients with both SI and RI overstimulation. This coincides with the observation that patients with ALS—in which both the SI and RI pathways are overstimulated—have a shorter life expectancy than patients with FTD, Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, and locked-in syndrome, diseases in which either SI or RI activity is increased.\n\n\nHomeostatic interactions between inhibitory transmission and excitatory transmission\n\nTaken together, the wealth of observations discussed above suggest that the opposing excitatory and inhibitory systems may play a role in the pathogenesis of the same disease. This phenomenon has precedent, as inhibitory/excitatory homeostasis processes are also involved in seizure activity30,31. Neurons that receive excessive excitatory stimulation can subsequently become overstimulated by inhibitory transmission, and vice versa. This raises the intriguing question of which system in neuromuscular and neurological diseases is overstimulated first, and which system becomes overstimulated as a homeostatic response. This question has been addressed with respect to epileptic seizures30. With respect to neuromuscular and neurological diseases, it is important to note that the administration of glutamatergic excitatory compounds does not lead to the clinical manifestations summarized in Table 1; glutamatergic overstimulation can give rise to clinical manifestations only through excitotoxicity (i.e., overstimulation-induced neuronal cell death). However, inhibitory overstimulation can give rise to the clinical manifestations in Table 1, even in the absence of neuronal cell death. Thus, I hypothesize that inhibitory overstimulation occurs first, and excitatory overstimulation is a homeostatic response. As inhibitory overstimulation increases, the excitatory system is stimulated further, until it reaches a level that induces neuronal cell death. This process is depicted schematically in Figure 2. Importantly, the order of the homeostatic process hypothesized here is precisely opposite to the homeostatic processes observed during epileptic seizures, in which excitatory overstimulation proceeds inhibitory overstimulation30.\n\nIn the inhibitory overstimulation hypothesis, excitatory overstimulation is a homeostatic response to inhibitory overstimulation. A key feature of this model is that inhibitory overstimulation can be sufficient to cause symptoms (left blue arrow). As the disease progresses, increasing inhibitory overstimulation can eventually lead to excitatory overstimulation and neuronal cell death, making the symptoms irreversible.\n\n\nOther possible interpretations of these observations\n\nOther possible interpretations of the findings summarized in Table 1 should be considered. If inhibitory overstimulation plays a key role in neuromuscular and neurological diseases, one would expect these patients to present with sedation, thus indicating the possibility that physiological systems other than the inhibitory system may be involved. However, not all GABA receptor subtypes are involved in sedation69,70. Thus, inhibitory activity can occur without inducing pronounced sedation. This is supported by reports that benzodiazepine-induced dysphagia can occur even in non-sedated patients38–40. Second, the absence of seizures despite high glutamate levels could be due to a slow, but non-epileptogenic increase in glutamate levels during the progression of neuromuscular and neurological diseases. However, even small increases in glutamate levels can increase glutamatergic synchronization of a small subset of critical neurons, thereby leading to epileptic activity30,31. Moreover, epileptic seizures are simply not reported among ALS patients, an observation that cannot be explained by a slow increase in glutamate levels, as glutamate does not cause inhibitory activity22–29. Furthermore, slow increasing levels of glutamate in the absence of seizure activity may reflect the involvement of an inhibitory homeostatic process30,31. Finally, the differentiation between SI and RI depicted in Table 4 and Table 5 may be attributed to the involvement of neuronal pathways projecting to either voluntary or involuntary muscles. However, this cannot explain split-hand syndrome in ALS patients or rest tremors in Parkinson’s disease, as these phenomena involve only voluntary muscles. Moreover, split-hand syndrome involves identically innervated muscles that cannot be differentiated by any aspect other than SI/RI innervation.\n\n\nTherapeutic potential for targeting inhibitory activity\n\nFrom a clinical perspective, an important consequence that emerges from the inhibitory overstimulation hypothesis is that the clinical manifestations summarized in Table 1 develop before neurons have undergone cell death. The implication of this possibility is that decreasing inhibitory activity may be beneficial in terms of slowing—or even preventing—the progression of neuromuscular and neurological diseases. Compounds that can reduce inhibitory activity are currently available; unfortunately, however, these compounds can induce seizure activity and are therefore not used therapeutically. Nevertheless, their potential for preventing the pathogenesis of neuromuscular and neurological diseases suggests that compounds that target the inhibitory system could be developed for clinical applications. For example, the average life expectancy of a patient with ALS is 3–5 years after onset, and most neuromuscular and neurological diseases are severe and ultimately fatal. With respect to Alzheimer’s disease and Parkinson’s disease, dysphagia and respiratory depression–related aspiration pneumonia are the most common causes of death16,17. Neuromuscular diseases also present with the severe and potential fatal clinical manifestations listed in Table 1;1–7,9–11,13,15,17–21 thus, the ability to prevent these symptoms could significantly prolong the life expectancy of these patients. From a treatment perspective, it is interesting to note that the selective GABA antagonist SGS-742 has been shown to be both clinically feasible and safe52.\n\n\nConclusions and future perspectives\n\nBased upon the plethora of observations regarding the inhibitory system, I hypothesize that this system plays an important role in the pathogenesis of both neuromuscular and neurological diseases. Importantly, overstimulation of the inhibitory system can explain both the absence of epileptic seizures despite the elevated glutamate levels and the pharmacological induction of symptoms present in patients with neuromuscular and neurological diseases. Moreover, the separation between SI and RI can account for the various categories of clinical manifestations observed in these patients. Specifically, I hypothesize that increased glutamate levels in neuromuscular and neurological diseases are actually a homeostatic response to an overstimulated inhibitory system. Implicating the inhibitory system in the pathogenesis of neuromuscular and neurological diseases is highly relevant, given that the majority of approaches being developed for treating these diseases focus on reducing glutamatergic activity, rather than reducing inhibitory activity. Moreover, this putative connection between the inhibitory system and neuromuscular/neurological diseases may have long-reaching implications, including the need to develop therapies designed to reduce inhibitory overstimulation in neuromuscular and neurological patients.",
"appendix": "Author contributions\n\n\n\nBT conceived the study, performed the analysis, and wrote the paper.\n\n\nCompeting interests\n\n\n\nBT has filed a patent application on the treatment of neuromuscular and neurologic diseases with therapies that reduce inhibitory overstimulation, and founded Ry Pharma, a company that aims to develop such therapies.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe author wishes to thank Dr. H. Jousma, Prof. Dr. A.F. Cohen and Prof. Dr. D.D. Breimer for their critical review of the manuscript.\n\n\nReferences\n\nStandifer: Handbook of disabilities, neuromuscular disorders, RCEP7, University of Missouri, 2014.\n\nKiernan MC, Vucic S, Cheah BC, et al.: Amyotrophic lateral sclerosis. Lancet. 2011; 377(9769): 942–55. PubMed Abstract | Publisher Full Text\n\nSilani V, Messina S, Poletti B, et al.: The diagnosis of Amyotrophic lateral sclerosis in 2010. Arch Ital Biol. 2011; 149(1): 5–27. PubMed Abstract | Publisher Full Text\n\nTurner MR, Hardiman O, Benatar M, et al.: Controversies and priorities in amyotrophic lateral sclerosis. Lancet Neurol. 2013; 12(3): 310–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAchi EY, Rudnicki SA: ALS and Frontotemporal Dysfunction: A Review. Neurol Res Int. 2012; 2012: 806306. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZago S, Poletti B, Morelli C, et al.: Amyotrophic lateral sclerosis and frontotemporal dementia (ALS-FTD). Arch Ital Biol. 2011; 149(1): 39–56. PubMed Abstract | Publisher Full Text\n\nRingholz GM, Appel SH, Bradshaw M, et al.: Prevalence and patterns of cognitive impairment in sporadic ALS. Neurology. 2005; 65(4): 586–90. PubMed Abstract | Publisher Full Text\n\nWarren JD, Rohrer JD, Rossor MN: Clinical review. Frontotemporal dementia. BMJ. 2013; 347: f4827. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTartaglia MC, Rowe A, Findlater K, et al.: Differentiation between primary lateral sclerosis and amyotrophic lateral sclerosis: examination of symptoms and signs at disease onset and during follow-up. Arch Neurol. 2007; 64(2): 232–6. PubMed Abstract | Publisher Full Text\n\nSinger MA, Statland JM, Wolfe GI, et al.: Primary lateral sclerosis. Muscle Nerve. 2007; 35(3): 291–302. PubMed Abstract | Publisher Full Text\n\nKalia LV, Lang AE: Parkinson's disease. Lancet. 2015; 386(9996): 896–912. PubMed Abstract | Publisher Full Text\n\nAlves L, Correia AS, Miguel R, et al.: Alzheimer’s disease: a clinical practice-oriented review. Front Neurol. 2012; 3: 63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoos RA: Huntington's disease: a clinical review. Orphanet J Rare Dis. 2010; 5: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpencer D: Seizures and epileptiform activity in early Alzheimer disease: how hard should we be looking? Epilepsy Curr. 2014; 14(2): 73–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartino D, Stamelou M, Bhatia KP: The differential diagnosis of Huntington's disease-like syndromes: 'red flags' for the clinician. J Neurol Neurosurg Psychiatry. 2013; 84(6): 650–656. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalia M: Dysphagia and aspiration pneumonia in patients with Alzheimer's disease. Metabolism. 2003; 52(10 Suppl 2): 36–8. PubMed Abstract | Publisher Full Text\n\nTjaden K: Speech and Swallowing in Parkinson’s Disease. Top Geriatr Rehabil. 2008; 24(2): 115–126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCohen B, Caroscio J: Eye movements in amyotrophic lateral sclerosis. J Neural Transm Suppl. 1983; 19: 305–15. PubMed Abstract\n\nPanicker JN, Fowler CJ, Kessler TM: Lower urinary tract dysfunction in the neurological patient: clinical assessment and management. Lancet Neurol. 2015; 14(7): 720–32. PubMed Abstract | Publisher Full Text\n\nToepfer M, Folwaczny C, Klauser A, et al.: Gastrointestinal dysfunction in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord. 1999; 1(1): 15–9. PubMed Abstract | Publisher Full Text\n\nFasano A, Visanji NP, Liu LW, et al.: Gastrointestinal dysfunction in Parkinson's disease. Lancet Neurol. 2015; 14(6): 625–39. PubMed Abstract | Publisher Full Text\n\nRothstein JD, Tsai G, Kuncl RW, et al.: Abnormal excitatory amino acid metabolism in amyotrophic lateral sclerosis. Ann Neurol. 1990; 28(1): 18–25. PubMed Abstract | Publisher Full Text\n\nSpreux-Varoquaux O, Bensimon G, Lacomblez L, et al.: Glutamate levels in cerebrospinal fluid in amyotrophic lateral sclerosis: a reappraisal using a new HPLC method with coulometric detection in a large cohort of patients. J Neurol Sci. 2002; 193(2): 73–78. PubMed Abstract | Publisher Full Text\n\nOlney JW: Neurotoxicity of excitatory amino acids. In: E.G. McGeer, J.W. Olney, P.L. McGeer (Eds.), Kainic Acid as a Tool in Neurobiology, Raven Press, New York, 1978; 95–121.\n\nCoyle JT, Puttfarcken P: Oxidative stress, glutamate, and neurodegenerative disorders. Science. 1993; 262(5134): 689–695. PubMed Abstract | Publisher Full Text\n\nKim AH, Kerchner GA, Choi DW: Blocking Excitotoxicity or Glutamatergic Storm. Chapter 1 in CNS Neuroprotection. Marcoux FW and Choi DW, editors. Springer, New York. 2002; 3–36. Reference Source\n\nLipton SA, Rosenberg PA: Excitatory amino acids as a final common pathway for neurologic disorders. N Engl J Med. 1994; 330(9): 613–622. PubMed Abstract | Publisher Full Text\n\nDoble A: The role of excitotoxicity in neurodegenerative disease: implications for therapy. Pharmacol Ther. 1999; 81(3): 163–221. PubMed Abstract | Publisher Full Text\n\nDong XX, Wang Y, Qin ZH: Molecular mechanisms of excitotoxicity and their relevance to pathogenesis of neurodegenerative diseases. Acta Pharmacol Sin. 2009; 30(4): 379–87. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuring MJ, Spencer DD: Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet. 1993; 341(8861): 1607–10. PubMed Abstract | Publisher Full Text\n\nChamberlin NL, Traub RD, Dingledine R: Role of EPSPs in initiation of spontaneous synchronized burst firing in rat hippocampal neurons bathed in high potassium. J Neurophysiol. 1990; 64(3): 1000–8. PubMed Abstract\n\nMöhler H, Crestani F, Rudolph U: GABAA-receptor subtypes: a new pharmacology. Curr Opin Pharmacol. 2001; 1(1): 22–5. PubMed Abstract | Publisher Full Text\n\nRudolph U, Knoflach F: Beyond classical benzodiazepines: Novel therapeutic potential of GABAA receptor subtypes. Nat Rev Drug Discov. 2011; 10(9): 685–697. PubMed Abstract | Publisher Full Text | Free Full Text\n\nForster A, Gardaz JP, Suter PM, et al.: Respiratory depression by midazolam and diazepam. Anesthesiology. 1980; 53(6): 494–7. PubMed Abstract | Publisher Full Text\n\nWall BT, Dirks ML, Snijders T, et al.: Substantial skeletal muscle loss occurs during only 5 days of disuse. Acta Physiol (Oxf). 2014; 210(3): 600–11. PubMed Abstract | Publisher Full Text\n\nSullivan EV, Pfefferbaum A: Alcohol and the Nervous System: Handbook of Clinical Neurology. Edited Edith V. Sullivan & Adolf Pfefferbaum. ISBN9780444626226. 2014; 95–96.\n\nHockman CH, Weerasuriya A, Bieger D: GABA receptor-mediated inhibition of reflex deglutition in the cat. Dysphagia. 1996; 11(3): 209–215. PubMed Abstract | Publisher Full Text\n\nWyllie E, Wyllie R, Cruse RP, et al.: The mechanism of nitrazepam-induced drooling and aspiration. N Engl J Med. 1986; 314(1): 35–8. PubMed Abstract | Publisher Full Text\n\nBuchholz D, Jones B, Neumann S, et al.: Benzodiazepine-induced pharyngeal dysphagia: a report of two probable cases. Dysphagia. 1995; 10(2): 142–6.\n\nDantas RO, Nobre Souza MA: Dysphagia induced by chronic ingestion of benzodiazepine. Am J Gastroenterol. 1997; 92(7): 1194–6. PubMed Abstract\n\nDua KS, Surapaneni SN, Santharam R, et al.: Effect of systemic alcohol and nicotine on airway protective reflexes. Am J Gastroenterol. 2009; 104(10): 2431–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Steveninck AL, Mandema JW, Tuk B, et al.: A comparison of the concentration-effect relationships of midazolam for EEG-derived parameters and saccadic peak velocity. Br J Clin Pharmacol. 1993; 36(2): 109–115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenson PJ, Beedie SA, Shephard E, et al.: Simple viewing tests can detect eye movement abnormalities that distinguish schizophrenia cases from controls with exceptional accuracy. Biol Psychiatry. 2012; 72(9): 716–24. PubMed Abstract | Publisher Full Text\n\nIyer PM, Egan C, Pinto-Grau M, et al.: Functional Connectivity Changes in Resting-State EEG as Potential Biomarker for Amyotrophic Lateral Sclerosis. PLoS One. 2015; 10(6): e0128682. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIgawa Y, Mattiasson A, Andersson KE: Effects of GABA-receptor stimulation and blockade on micturition in normal rats and rats with bladder outflow obstruction. J Urol. 1993; 150(2 Pt 1): 537–42. PubMed Abstract\n\nAndersson KE, Wein AJ: Pharmacology of the lower urinary tract: basis for current and future treatments of urinary incontinence. Pharmacol Rev. 2004; 56(4): 581–631. PubMed Abstract | Publisher Full Text\n\nHyland NP, Cryan JF: A Gut Feeling about GABA: Focus on GABAB Receptors. Front Pharmacol. 2010; 1: 124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrantis A: GABA in the Mammalian Enteric Nervous System. News Physiol Sci. 2000; 15: 284–290. PubMed Abstract\n\nOslin D, Atkinson RM, Smith DM, et al.: Alcohol related dementia: proposed clinical criteria. Int J Geriatr Psychiatry. 1998; 13(4): 203–212. PubMed Abstract | Publisher Full Text\n\nBillioti de Gage S, Bégaud B, Bazin F, et al.: Benzodiazepine use and risk of dementia: prospective population based study. BMJ. 2012; 345(2012): e6231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBell GD, Spickett GP, Reeve PA, et al.: Intravenous midazolam for upper gastrointestinal endoscopy: a study of 800 consecutive cases relating dose to age and sex of patient. Br J Clin Pharmacol. 1987; 23(2): 241–243. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFroestl W, Gallagher M, Jenkins H, et al.: SGS742: the first GABAB receptor antagonist in clinical trials. Biochem Pharmacol. 2004; 68(8): 1479–87. PubMed Abstract | Publisher Full Text\n\nMartínez-Cué C, Delatour B, Potier MC: Treating enhanced GABAergic inhibition in Down syndrome: Use of GABA α5-selective inverse agonists. Neurosci Biobehav Rev. 2014; 46(Pt 2): 218–227. PubMed Abstract | Publisher Full Text\n\nWu Z, Guo Z, Gearing M, et al.: Tonic inhibition in dentate gyrus impairs long-term potentiation and memory in an Alzheimer's [corrected] disease model. Nat Commun. 2014; 5: 4159. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiller RG, Moore DH 2nd, Gelinas DF, et al.: Phase III randomized trial of gabapentin in patients with amyotrophic lateral sclerosis. Neurology. 2001; 56(7): 843–848. PubMed Abstract\n\nCudkowicz ME, Shefner JM, Schoenfeld DA, et al.: A randomized, placebo-controlled trial of topiramate in amyotrophic lateral sclerosis. Neurology. 2003; 61(4): 456–464. PubMed Abstract | Publisher Full Text\n\nGordon PH, Moore DH, Miller RG, et al.: Efficacy of minocycline in patients with amyotrophic lateral sclerosis: a phase III randomised trial. Lancet Neurol. 2007; 6(12): 1045–53. PubMed Abstract | Publisher Full Text\n\nAhmadirad N, Shojaei A, Javan M, et al.: Effect of minocycline on pentylenetetrazol-induced chemical kindled seizures in mice. Neurol Sci. 2014; 35(4): 571–6. PubMed Abstract | Publisher Full Text\n\nElia AE, Lalli S, Monsurrò MR, et al.: Tauroursodeoxycholic acid in the treatment of patients with amyotrophic lateral sclerosis. Eur J Neurol. 2016; 23(1): 45–52. PubMed Abstract | Publisher Full Text\n\nYanovsky Y, Schubring SR, Yao Q, et al.: Waking action of ursodeoxycholic acid (UDCA) involves histamine and GABAA receptor block. PLoS One. 2012; 7(8): e42512. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWindhorst U: On the role of recurrent inhibitory feedback in motor control. Prog Neurobiol. 1996; 49(6): 517–87. PubMed Abstract | Publisher Full Text\n\nLipski J, Fyffe RE, Jodkowski J: Recurrent inhibition of cat phrenic motoneurons. J Neurosci. 1985; 5(6): 1545–55. PubMed Abstract\n\nMcCoy T: How Stephen Hawking, diagnosed with ALS decades ago, is still alive. Washington Post, February 24, 2015. accessed May 10, 2016. Reference Source\n\nEisen A, Kuwabara S: The split hand syndrome in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry. 2012; 83(4): 399–403. PubMed Abstract | Publisher Full Text\n\nNapier JR: The prehensile movements of the human hand. J Bone Joint Surg Br. 1956; 38-B(4): 902–913. PubMed Abstract\n\nRossi A, Mazzocchio R: Renshaw recurrent inhibition to motoneurones innervating proximal and distal muscles of the human upper and lower limbs. In: Muscle Afferents and Spinal Control of Movement, Eds L. Jami, E. Pierrot-DeseilIigny and D Zytnicki. Pergamon Press: Oxford. 1992; 313–319.\n\nHelmich RC, Hallett M, Deuschl G, et al.: Cerebral causes and consequences of parkinsonian resting tremor: a tale of two circuits? Brain. 2012; 135(Pt 11): 3206–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmabeoku GJ, Chikuni O: Effects of some GABAergic agents on quinine-induced seizures in mice. Experientia. 1992; 48(7): 659–62. PubMed Abstract | Publisher Full Text\n\nMcKernan RM, Rosahl TW, Reynolds DS, et al.: Sedative but not anxiolytic properties of benzodiazepines are mediated by the GABA(A) receptor alpha1 subtype. Nat Neurosci. 2000; 3(6): 587–92. PubMed Abstract | Publisher Full Text\n\nReynolds DS, Rosahl TW, Cirone J, et al.: Sedation and anesthesia mediated by distinct GABA(A) receptor isoforms. J Neurosci. 2003; 23(24): 8608–17. PubMed Abstract"
}
|
[
{
"id": "15214",
"date": "05 Aug 2016",
"name": "Janet Hoogstraate",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript by B.Tuk provides a comprehensive and up-to-date review of the clinical observations of neurodegenerative diseases relevant to the discussion on pathogenesis. The methodology of linking the clinical observations to a common pathology that is different in sequence of events compared to other scientific literature is sound. The data and the proposed theory are presented in a suitable manner. The position that there is an absence of elevated seizure activity in all 8 of the neurological and neuromuscular diseases should be revised to my opinion. While this can be said for ALS, there is substantial epidemiological data to support an higher frequency of epileptic seizures in the early stages of Alzheimer's Disease (ref 1-5). Based on the enormous increase of glutamate levels in ALS patients and the absence of elevated risk for epileptic activity in that population, there is still good support for postulating the theory of the inhibitory system as a plausible trigger for these diseases, but I propose that that section of the article and Table 1 be revised to reflect that nuance.\nI recommend that the article be indexed, with the minor revision outlined above.",
"responses": []
},
{
"id": "15213",
"date": "10 Aug 2016",
"name": "Willem M de Vos",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe opinion paper by Tuk provides an overview of the inhibitory system involved in neural signal transduction. Based on an extensive review of the literature the author provides a number of challenging but well-reasoned hypotheses about the impact of this system on both neuromuscular and neurological diseases. This potential connection may have important implications for the understanding and future treatment of these so far (almost) untreatable diseases.\n\nOnly few minor points should be addressed.\n\nThe use of the term ‘inhibitory system’ is somewhat confusing – the context is often not specified (inhibitory system of what…) although the author means neural signal transduction – however, this should be mentioned where relevant and certainly in the title.\n\nThe link between GABA production by intestinal bacteria and gut GABA receptor stimulation could be further addressed as it provides a way how other factors such as food or the gut microbiome may modulate GABA response (see also reference 47 and later work of these authors like Bravo et al. (2011).\n\nThe literature is well reviewed and the paper is supported by 5 useful Tables, 2 clear Figures and 70 references of peer reviewed papers – the one Washington Post article may be better mentioned in the text.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1435
|
https://f1000research.com/articles/5-1528/v1
|
28 Jun 16
|
{
"type": "Observation Article",
"title": "Neck keloids: evaluation of risk factors and recommendation for keloid staging system",
"authors": [
"Michael H. Tirgan"
],
"abstract": "Importance: Health care providers have long struggled with recurrent and hard to treat keloids.\n\nAdvancing our understanding of natural history and risk factors for development of large, very large and massive neck keloids can lead to improved treatment outcomes. Clinical staging system for the categorization of keloid lesions, as well as grouping of keloid patients according to the extent of skin involvement is both fundamental for design and delivery of proper plan of care and an absolute necessity for methodical trial design and interpretation of the results thereof. Objective: To review clinical presentation and natural history of neck keloids; to explore risk factors for development of large, very large and massive neck keloids; and to propose a clinical staging system that allows for categorization of keloid lesions by their size and grouping of keloid patients by the extent of their skin involvement. Setting: This is a retrospective analysis of 68 consecutive patients with neck keloids who were seen by the author in his keloid specialty medical practice.\n\nIntervention: Non-surgical treatment was offered to all patients. Results: Neck-area keloids were found to have several unique characteristics. All 53 African Americans in this study had keloidal lesions elsewhere on their skin. Very large and massive neck keloids appear to be race-specific and almost exclusively seen among African Americans.\n\nKeloid removal surgery was found to be the main risk factor for development of very large and massive neck keloids. Conclusions and relevance: Surgical removal of neck keloids results in wounding of the skin and triggering a pathological wound-healing response that often leads to formation of a much larger keloid. Given the potential for greater harm from surgery, the author proposes non-surgical approach for treatment of all primary neck keloids. Author’s attempts to properly categorize keloid lesions and to group the study subjects was hampered by the lack of a previously defined methodology. A clinical staging system is proposed to address the deficiency in grouping of keloid patients according to the size and extent of skin involvement with keloid lesions.",
"keywords": [
"Keloid disorder",
"Neck keloids",
"Keloid staging"
],
"content": "Introduction\n\nNeck-area keloids are fairly uncommon and seen most often among individuals with black skin. Inflammation and wounding of the skin, either from ingrown hair or shaving blades, are perhaps the leading triggering factors in formation of these keloids in genetically susceptible individuals. Although there are several publications about neck keloids, there is a void in medical literature about natural history or risk factors for development of large, very large and massive keloids in this anatomical region.\n\nKeloid Disorder (KD) is an inherited ailment of wound-healing processes1 characterized by highly variable clinical presentation that spans from individuals with one or very few small keloidal lesions to those with numerous and very large lesions involving large portions of their skin. To the author’s knowledge, no clinical staging system has been proposed to categorize KD patients according to the extent of their skin involvement.\n\nGenetics of KD remains poorly understood. However, clinical observation suggests that the genetic predisposition to KD has a wide spectrum, from individuals who suffer from mild form of the disorder who in their lifetime only develop one or few slow-growing keloidal lesions, to those with very severe form of the disorder and who develop numerous large and fast-growing keloids; and of course, there are many others who fall somewhere in between these two extremes. As with most other genetic illnesses, there also exist many individuals who are simply carriers of the gene, who may never become symptomatic.\n\nIn addition to the genetics, several other factors play critical roles in clinical presentation of KD. Most importantly, there must exist an injury to the skin that would trigger an abnormal wound-healing response which leads to formation of keloidal lesions. Obviously, there is a wide spectrum to the severity and extent of skin injuries, ranging from very minor insults to the skin —from acne, piercing, or vaccination— to more severe forms of skin injury from surgery or burns. Besides genetics and skin injury, other important factors are age, race, gender, chronicity, therapeutic interventions and location of the keloidal lesions. The wide spectrum of all these factors contributes to highly variable phenotypes of KD.\n\nAnatomically, the superior margin of the neck is, posteriorly at the level of superior nuchal line of the cranium and anteriorly at the level of the lower margin of the mandible. The inferior boundary of the neck is at the level of the suprasternal notch, the clavicle and the first rib. The skin boundaries between face, neck, and chest are irregular and hard to precisely define. For purpose of this publication and in accordance with the anatomical definitions, the keloids of the submandibular skin are grouped with other neck keloids.\n\n\nMaterials and methods\n\nThis is a retrospective analysis of 68 consecutive patients with neck keloids who were seen by the author in his keloid specialty medical practice. Fifteen patients were Caucasians or Asians, and 53 patients were African Americans. Patients were grouped according to the size of their neck keloids, regardless of presence or absence of keloid lesions elsewhere. Keloidal lesions were assessed visually and divided into four categories. Table 1 summarizes characteristics of the patients within each group.\n\n1. Massive neck keloids that were very bulky and extended to both sides of the neck, often involving much of the submandibular space, or covering a very large surface of neck. Keloidal lesions of 14 patients met this criteria. Thirteen of these patients were African Americans and one was Caucasian. Figure 1 depicts some of the patients in this group.\n\n2. Very large neck keloids, whereby the bulk of keloidal mass was limited to one side of the neck. Keloidal lesions of 10 patients met this criteria. Nine patients were African Americans and one was Caucasian. Figure 2 depicts some of patients in this group.\n\n3. Large neck keloids, whereby the keloid mass formed one solitary tumor, or was comprised of multiple small lesion. Keloidal lesions of 11 patients met this criteria. Ten of these patients were African Americans and one was Caucasian. Figure 3 depicts some of the patients in this group.\n\n4. Minimal neck keloids, whereby keloidal lesions were small in size, liner, or nodular but without formation of keloid tumors. Keloidal lesions of 33 patients met this criteria. Twenty-one patients were African Americans and 12 were Asian/Caucasian. Figure 4 depicts several of Asian or Caucasian patients. Figure 5 depicts some of the African American patients in this group.\n\nGreen dots represent two cases of de novo massive keloids. All other patients have previously undergone at least one keloid-removal surgery.\n\nGreen dot represents one case of de novo massive keloids. All other patients have previously undergone at least one keloid-removal surgery.\n\nThe author was unable to find a previously described methodology, or staging system, that would allow for more precise grouping of patients with neck or other keloids. Although volumetric ultrasound measurements can be objectively applied to skin lesions2, the author used visual inspection of the lesions as his sole method of grouping keloidal lesions, knowing that this method is subjective.\n\nThe IRB at Rockefeller University Hospital determined that publishing this work to be exempt from IRB review and approval process.\n\n\nResults\n\nDespite the fact that this retrospective analysis is limited by its small sample size, several observations were made with respect to the neck-area keloids.\n\n1. African Americans who present with neck keloids, often have keloidal lesions elsewhere in their skin. All 53 African Americans in this study had keloidal lesions elsewhere on their skin. This association was not as strong among 15 Asians and Caucasian patients, of these 11 patients had keloids elsewhere, and four patients had keloids only in their neck.\n\n2. Neck keloidal lesions among Asians and Caucasians are usually small in size and often nodular or linear; they rarely take on a tumoral form. Among all Asians and Caucasians in this study, only one patient developed massive submandibular keloid subsequent to numerus surgeries he had for removal of face and neck keloidal lesions.\n\n3. Keloid removal surgery as a mean of treatment for primary neck keloids is a clear risk factor for development of very large and massive secondary neck keloids. Among 24 patients who had massive or very large neck keloids, 22 two had previously undergone at least one prior keloid removal surgery.\n\nVery large and massive neck keloids are race specific. There were only two Caucasians among the 24 patients with massive and very large neck keloids. Summary of this data is presented in Table 2.\n\n\nDiscussion\n\nAlthough there are numerous publications about keloids, in particular focusing on ear keloids, there is paucity of literature about neck keloids. With exception of several case reports within body of general keloid publications, there is lack of authoritative guidance, randomized studies, or even expert opinions on the proper management of neck-area keloids. PubMed search using “neck” and “keloid” as two key words does not yield any results.\n\nThe skin of neck is an uncommon place for development of keloids. Indeed many KD patients, even those with the most severe form of KD, may never develop keloids in the neck-area. Figure 6 depicts four non-African American patients with moderate to severe forms of chest keloid, yet the skin of the neck in these cases seems to be spared from involvement by KD.\n\nBy far, the most important factor in the development of a primary keloidal lesion is the injury to skin that leads to triggering of a pathological wound-healing response. Although ear piercing is a well-recognized triggering factor for development of primary ear keloids, no such factor has been definitively implicated in neck keloids. Neck-area surgery is a known but uncommon triggering factor for formation of primary neck keloids3.\n\nNotice absence of neck keloids in these patients.\n\nKnowing that KD is a genetic disorder of wound-healing processes, it is counterintuitive to resort to surgery as the mainstay of treatment. Surgical removal of neck keloids is an intervention that is commonly practiced, not only by ear-nose-throat specialists, but also by plastic surgeons and general dermatologists. Surgical intervention however, defies the very basic principal in keloid formation. The injury and insult from surgery to the skin that surrounds a keloidal lesion, on its own, will undoubtedly trigger a keloidal wound-healing response that often leads to formation of a new keloid. Adjuvant treatments in form of post-operative steroid injections4 or even radiation therapy5 are commonly incorporated in management of every KD patient who undergoes surgery simply to counter the fully expected recurrence after surgery. Yet despite diligent use of all available adjuvant methods, a significant number of keloid patients undergo a second, third, or fourth surgery. In many unfortunate instances, keloids keep relapsing and at some point, either the surgeon or the patient —or both— gives up. The unfortunate patient ends up accepting the truth about inability of surgery to treat his or her keloids and sees no other choice but to surrender to living with huge tumoral keloids on his or her neck. Figure 1 and Figure 2 depict many such patients.\n\n\nStaging of keloid disorder\n\nAs is evident in the assessment of the size of keloid lesions and the grouping of subjects in this study, currently there is no staging system that would allow for proper categorization of keloidal lesions and meaningful grouping of keloid patients. This retrospective study is clearly handicapped by the lack of such a staging system.\n\nTNM cancer staging system has been used for several decades and allows proper stage grouping of cancer patients6. A great majority of oncology interventions, clinical trials, and standard treatments are guided by the TNM staging of the cancer at any given time. Conduct and interpretation of the result of oncology clinical trials are virtually dependent on this staging system.\n\nWithout such a staging system for KD, the interpretation of published study results is very difficult. For example, when a study looks at the rate of recurrence of ear keloids after surgery, common sense tells us that patients who only have one keloid on one ear, may have a different rate of recurrence from those who have numerous keloids on their skin. Also, those who have had prior surgical removal of their keloids will have a much higher rate of recurrence than those patients who have never had surgery before.\n\nTo assess each keloid patient properly, to better understand the natural history of this disorder, and to be able to compare future study results among various keloid patients groups, we clearly need a staging system that can allow us to describe the severity KD based on the size, location and/or extent of the keloidal lesions as well as history of surgery or radiation therapy, and perhaps other factors that are currently unknown to us. It is quite conceivable that proper management of KD patients could be guided by such a staging system. Clinical staging of KD would also be the only method that could stratify for such preexisting inherent risks of recurrence, such as response to prior treatments, positive family history, age, gender, race, and so forth.\n\nA well-designed clinical staging system will need to be validated by review of retrospective studies as well as planned prospective clinical trials. The author hereby proposes the following staging system for KD patients.\n\n\nClinical staging and classification of keloids\n\nStage 0: Genetically predisposed. At least one parent has had keloids. Index person has no clinical evidence or history of keloid or any hypertrophic scars.\n\nStage I: Presence of only one keloidal lesion.\n\nStage 1A: Presence of only one keloidal lesion that measures no greater than 2 centimeters in any dimension.\n\nStage 1B: Presence of only one keloidal lesion that measures 2.1 – 10 centimeters in any dimension.\n\nStage 1C: Presence of only one keloidal lesion that measures greater than 10 centimeters in any dimension.\n\nStage II: Presence of multiple keloidal lesion. The sum of the largest diameter of the keloids is up to 30 centimeters.\n\nStage II A: Keloids measure ≤ 2 centimeters in largest diameter; the sum of the largest diameter of all keloids measures 10 centimeters or less.\n\nStage II B: Keloids measure ≤ 10 centimeters in largest diameter, at least one keloid measures 2.1 – 10 centimeters in its largest diameter; the sum of the largest diameter of all keloids measures 10.1 – 20 centimeters.\n\nStage II C: At least one keloid measures 10 centimeters in largest diameter; the sum of the largest diameter of all keloids measures up to 30 centimeters.\n\nStage III: Presence of multiple keloidal lesions; the sum of the largest diameter of the keloids measures 30.1 – 50 centimeters.\n\nStage III A: Keloids measure ≤ 2 centimeters in largest diameter; the sum of the largest diameter of all keloids measure 30.1 – 40 centimeters.\n\nStage III B: Keloids measure ≤ 10 centimeters in largest diameter; at least one keloid measures 2.1 – 10 centimeters in its largest diameter; the sum of the largest diameter of all keloids measures 30.1 – 40 centimeters.\n\nStage III C: At least one keloid measures 10 centimeters in largest diameter; the sum of the largest diameter of all keloids measures 30.1 – 50 centimeters.\n\nStage IV: Presence of multiple keloidal lesions; the sum of the largest diameter of the keloids is greater than 50 centimeters.\n\nStage IV A: Keloids measure ≤ 2 centimeters in largest diameter; the sum of the largest diameter of all keloids measures greater than 50 centimeters.\n\nStage IV B: Keloids measure ≤ 10 centimeters in largest diameter; at least one keloid measures 2.1 – 10 centimeters in its largest diameter. The sum of the largest diameter of all keloids measures greater than 50 centimeters.\n\nStage IV C: At least one keloid measures greater than 10 centimeters in its largest diameter; the sum of the largest diameter of all keloids to measure greater than 50 centimeters.\n\nTable 3 is a summary of stage grouping of patients within each group.\n\n“SYMP” will designate a keloid as symptomatic. Symptoms can include pain, itching, bleeding, infection and so forth. “SURG” designation indicates to prior history of surgery for the index keloidal lesion(s). The designation “n” indicates to the number of prior surgical attempts to remove a keloid. “RAD” designation indicates to a prior history of radiation therapy for the any keloidal lesion(s) in a particular patient.\n\n\nConclusions\n\nUnderstanding the natural history of KD and recognition of risk factors that lead to formation of large, very large and massive neck keloids are the most important and fundamental elements for development of safe and effective treatment strategies. The goal of treatment for all KD patients, and those with neck-area keloids in particular, should pivot not only on removal of the keloid tissue but most importantly on prevention of the recurrence of the keloid. Performing surgery to remove primary keloidal lesions is inherently contrary to both these principles. Surgery by its nature induces a totally new injury to the skin and triggers the same dysregulated wound-healing response to a new and more extensive dermal injury which is the causal factor for formation of very large and massive neck keloids.\n\nSystematic use of non-surgical interventions as a primary mode of treating all neck-area keloids by all health care providers will most certainly prevent the development of large, very large and incurable massive keloids. This approach will also eliminate the need for hazardous adjuvant radiation therapy.\n\nFurthermore, a staging system for categorizing keloids and grouping of patients according to the specifics of their KD lesions will better identify the patients’ keloid-specific characteristics. Clinical staging system also has the potential to become the foundation for design and delivery of individualized plan of care. It can also become the core for methodical trial design and accurate interpretation of the study findings.\n\n\nData availability\n\nAll raw data relevant to the study are provided in tables above.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author declares that no grants were involved in supporting this work.\n\n\nReferences\n\nMarneros AG, Norris JE, Olsen BR, et al.: Clinical genetics of familial keloids. Arch Dermatol. 2001; 137(11): 1429–34. PubMed Abstract | Publisher Full Text\n\nMandava A, Ravuri PR, Konathan R: High-resolution ultrasound imaging of cutaneous lesions. Indian J Radiol Imaging. 2013; 23(3): 269–277. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYoung WG, Worsham MJ, Joseph CL, et al.: Incidence of keloid and risk factors following head and neck surgery. JAMA Facial Plast Surg. 2014; 16(5): 379–80. PubMed Abstract | Publisher Full Text\n\nShons AR, Press BH: The treatment of earlobe keloids by surgical excision and postoperative triamcinolone injection. Ann Plast Surg. 1983; 10(6): 480–482. PubMed Abstract\n\nvan Leeuwen MC, Stokmans SC, Bulstra AE: Surgical Excision with Adjuvant Irradiation for Treatment of Keloid Scars: A Systematic Review. Plast Reconstr Surg Glob Open. 2015; 3(7): e440. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSobin LH: TNM: evolution and relation to other prognostic factors. Semin Surg Oncol. 2003; 21(1): 3–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14892",
"date": "18 Jul 2016",
"name": "Lamont R. Jones",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe author has attempted to address a problem that has hindered the ability to compare research and treatment outcomes for keloids. Besides a universally agreed upon marker for keloids, a staging system is paramount to successfully compare studies on keloid pathogenesis and treatment outcomes. There are some limitations to this article that should be addressed prior to publication.\nPlacing Caucasian and Asian patients into one group, although it may have been necessary in this study due to small ample size, is not optimum because of differences in risks for keloid development and treatment.\n\nThe grouping of keloids into massive, very large, large, and minimal neck keloids seems arbitrary and the clinical significance is unknown.\n\nI agree that the conclusion that surgery for neck keloids should not be the first choice, especially for large keloids. However, information on the size of the keloids prior to the initial surgical excision is lacking and would add more validity. For example, if the keloid was 3 cm prior to initial excision and 6 cm afterwards compared to 6 cm prior to surgery and 6 cm afterwards. Furthermore, information on outcomes for non-surgical treatment options is needed.\n\nStaging systems are intended to guide treatment and to predict outcomes. As presented the staging system does not satisfy either purpose. The staging system is only descriptive and its clinical implications are unknown. Larger number are needed to validate or refute the proposed staging system.",
"responses": [
{
"c_id": "2141",
"date": "19 Aug 2016",
"name": "Michael Tirgan",
"role": "Author Response",
"response": "Thank you for taking time to review and comment on my publication. I truly appreciate your detailed review, and each and every comment you have made. I hereby address the points you have raised. As for placing Asian and Caucasian patients in the same group as other patients, it was done for the following reasons: a) In all other diseases such as lung cancer, diabetes, hypertension, we do include all patients, yet we are cognizant of the biological differences that do exist among various subgroups of patients. This manuscript reviews neck area keloids, therefore all patients with neck keloids, regardless of their race, heritage or gender were included. b) I attempted to show the very interesting differences in the morphology of the disorder among races, hoping that one day we can understand the biological differences that result in a clear difference in phenotype of the disorder. As for grouping of patients in four different categories, it was done to lay the foundation that would support the proposed staging system. As you are well aware, keloid disorder is perhaps the most under-studied human ailment. We even struggle with proper terminology for it, as some still call it “keloid scarring”. As documented by numerous images, we are dealing with a semi-neoplastic process that can lead to formation of massive tumors, and yet in 2016 we have had no methods to classify, or group these patients. Therefore, the study patients were grouped according to the size of their neck keloid lesions, a methodology that is commonly used in staging tumors according to the size of the primary tumors using TNM staging system. As for the size of keloid lesions, and medical treatment outcomes, both your point are well taken as I did not report them. The focus of this manuscript was to simply classify neck keloids and show the absolute need for a staging system. The conclusions drawn, as to surgery being the dominant risk factor for massive and very large keloids, is simply supported by the data that was presented. I would love to learn about your data as to the occurrence of de novo massive and very large neck keloids. As for staging system, I agree with all your comments. I hope that the staging system that is proposed here, however inadequate it may be, to be a starting point, so that we can all work towards improving it, or coming up with an even better one. The undisputed fact remains that we have never had a staging system for keloid disorder. Thank you again for all your comments."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1528
|
https://f1000research.com/articles/5-2032/v1
|
18 Aug 16
|
{
"type": "Case Report",
"title": "Case Report: Post transcatheter aortic valve replacement shock: Value of multimodal imaging",
"authors": [
"Sujatha P. Bhandary",
"Andrew J. Otey",
"Thomas J. Papadimos",
"Juan A. Crestanello",
"Barry S. George",
"Katja R. Turner",
"Andrew J. Otey",
"Thomas J. Papadimos",
"Juan A. Crestanello",
"Barry S. George",
"Katja R. Turner"
],
"abstract": "Complications resulting from the delayed clinical presentation of a left main coronary artery obstruction can be catastrophic. This case report presents a 73-year-old woman with severe aortic stenosis who underwent transcatheter aortic valve replacement with a core valve who, approximately 20 minutes after heparin reversal with protamine, became hypotensive and was unresponsive to vasopressor and inotropic therapy. Transesophageal echocardiography demonstrated global hypokinesis, which was highly consistent with the occlusion of the left main coronary artery. Angiography confirmed this diagnosis and demonstrated that valve positioning had not changed compared to post-placement examination. Here we report the partial covering of the ostium of the left main coronary artery by a core valve skirt that converted into a total occlusion following the initiation of heparin reversal with protamine and the value of multimodal imaging in the management of this case.",
"keywords": [
"Aortic stenosis",
"transcatheter aortic valve replacement",
"complications",
"intraoperative transesophageal echocardiography",
"coronary occlusion"
],
"content": "Introduction\n\nTranscatheter aortic valve replacement (TAVR) is an increasingly popular approach for high-risk patients with severe aortic stenosis. Coronary ostium occlusion after valve implantation is a life-threatening complication that occurs in up to 0.4% of TAVR procedures despite the growing experience in performing this procedure and the constant improvements in TAVR devices1,2. We report the partial occlusion of the ostium of the left coronary artery (LCA) by the skirt of the core valve that was converted into a total occlusion following the initiation of heparin reversal with protamine and the value of multimodal imaging in the management of this case.\n\n\nCase report\n\nA 73-year-old woman with severe aortic stenosis (aortic valve area 0.7 cm2 and a mean gradient of 60 mmHg), preserved left ventricular (LV) systolic function, left ventricular hypertrophy, no significant coronary artery disease, and an Society of Thoracic Surgeons (STS) score of 3 presented to our institution for an aortic valve replacement (http://tools.acc.org/TAVRRisk/). The patient had compromised pulmonary function (FEV1 46% and DLCO 46%), and a catheter-based valve replacement was performed. Preoperative computer tomographic derived measurements demonstrated: aortic valve annulus diameter of 25 mm (major) and 20 mm (minor); coronary artery ostium distance of 18 mm (right) and 11.6 mm (left); coronary sinus diameter/height of 26/18 mm (right coronary), 28/19 mm (left coronary), 28/19 mm (non-coronary); and a moderate severe calcified valve with moderate-severe impaired leaflet excursion (Figure 1). A 26 mm Medtronic CoreValve (Medtronic, Minneapolis, MN, USA) was implanted via a right ileo-femoral approach. The patient had an uneventful valve deployment. A preoperative transesophageal echocardiography (TEE) verified a calcific, severely stenotic aortic valve with a normal left ventricular systolic function (ejection fraction 60% (Video 1 & Video 2). Deployment was facilitated using brief, temporary rapid pacing at 120 beats per minute; the patient was hemodynamically stable after valve deployment. A subsequent TEE examination revealed mild aortic paravalvular regurgitation (Video 3). A post-deployment angiography demonstrated a cephaled migration of the valve (from position 2 to -2), but coronary artery perfusion was maintained, as shown in a right anterior oblique (RAO) image (Figure 2B & Figure 3). Based on the images provided and the hemodynamic stability of the patient, the operative team was satisfied with the position of the valve and focused on repairing the femoral access site. The access site was repaired with normal flow twenty minutes after deployment as demonstrated by angiography. The decision was then made to reverse the anticoagulation with protamine. Shortly after the slow administration of protamine, the patient became hypotensive and unresponsive to vasopressor or inotropic therapy.\n\nPreoperative measurements of the aortic valve.\n\nA. Details post-deployment of core valve. B. RAO view of the deployed valve, detailing flow. C. LAO view depicting lack of flow.\n\nA TEE probe was emergently placed for examination and revealed normal right ventricular size and function, with global hypokinesis of the left ventricle (Video 4). These findings were consistent with a left main coronary artery occlusion in a left dominant system patient. Cardiopulmonary bypass (CPB) was instituted as a result of hemodynamic instability, and the diagnosis of coronary occlusion was confirmed with angiography (Figure 4). Due to the position of the valve, the decision was made to refrain from coronary stent placement and to convert to an open surgical replacement of the aortic valve. After the Medtronic CoreValve was removed, the left coronary ostium was examined and determined to be patent. After the placement of a 21 mm Perimount Magna valve (Edwards Lifesciences, Irvine, CA, USA), the patient was weaned from CPB without any problems. Further TEE was performed which revealed a normal left ventricular function. The patient was extubated on postoperative day one and remained hemodynamically stable without further concerns regarding her aortic valve or coronary ischemia. On the 7th postoperative day, the patient developed complete heart block that necessitated the placement of a dual chamber permanent pacemaker. She was discharged on postoperative day eight.\n\nRAO coronary angiography showing flow through the left main coronary artery and branches.\n\n\n\n\n\n\n\n\n\n\nDiscussion\n\nTAVR is a novel approach for the treatment of high-risk patients with severe aortic stenosis, especially those deemed inoperable. Complications associated with TAVR include: (1) access related problems, including bleeding and occlusion (dissection or vascular); and (2) procedure-related problems such as cardiac tamponade, root rupture (annular or aortic), apical tears, right ventricular perforation, aortic regurgitation, coronary artery occlusion, renal failure, stroke, and conduction system disturbances requiring permanent pacemaker implantation.\n\nLeft main coronary artery occlusion after TAVR is a rare but potentially fatal complication. A complete occlusion of the left main coronary artery will manifest itself immediately; however, a partial compromise can be initially silent, such as the occlusion in our case. A partial LCA occlusion can result if calcific particles from the native aortic leaflet become displaced over the ostium while the aortic valve device is being expanded3,4. Other proposed mechanisms include incorrect positioning of the valve frame (obstructive portion) directly over the coronary ostium, hematoma formation or leaflet distortion, and apposition on the LCA ostium5,6. Valve design is an additional factor related to coronary obstruction7. Higher probabilities of occlusions occur with the balloon-expandable SAPIEN XT valve (Edwards Lifesciences, Irvine, CA, USA) than the Medtronic CoreValve4. This can be attributed to the hourglass-like design of the Medtronic CoreValve, which is less prone to native cusps displacement over the ostia4. However, a previous case has been reported using the Medtronic CoreValve that resulted in a left main coronary occlusion from calcium nodules3.\n\nIn our case, hemodynamic instability developed shortly after the administration of protamine. The initial working diagnosis suggested this instability resulted from an adverse reaction to the drug. Cardiovascular reactions to protamine vary from vasodilation-related hypotension to pulmonary hypertension, with subsequent right heart failure and cardiovascular collapse8. While treating our patient’s hypotension, which was unresponsive to vasopressors and inotropes, a TEE probe was placed and demonstrated global LV dysfunction. This observation was consistent with a LCA occlusion. Since neither right ventricular (RV) dysfunction nor indirect evidence of pulmonary hypertension were present, causality related to a severe protamine reaction was not likely.\n\nIn our case, a post-deployment coronary angiogram was done in the RAO projection and showed adequate coronary blood flow (Figure 3). Following the emergent implementation of CPB, another coronary angiography was performed in the left anterior oblique (LAO) projection and demonstrated complete cessation of flow through the left coronary ostium (Figure 4). Due to the high position (-2) of the Medtronic CoreValve, it was deemed impossible to place a stent across the occluded ostium and restore coronary blood flow in a timely fashion without lasting sequelae. Thus, the decision was made to surgically replace the calcific native valve after removing the prosthesis in place. After visualization of the valve in situ, it was noted that the skirt of the valve partially covered the left main coronary ostium. Due to the slightly high position of the prosthetic valve and reduced flow in the native “sinotubular reservoir”, protamine administration may have facilitated the formation of small micro thrombi aggregates adherent to the prosthetic material and led to a complete coronary occlusion (Figure 5). Even though a post-deployment cephalad migration of the valve (from 2 to -2) was immediately noted and raised concern regarding a coronary occlusion, an angiogram in the ROA view demonstrated good coronary flow. The assumption was made that the valve was situated in a fashion where adequate blood flow was maintained through the cage portion of the valve and was unobstructed by the overlying curtain (Figure 3).\n\nLAO coronary angiography showing lack of perfusion and stasis of blood within the left sinus of Valsalva.\n\nThis is the image of the core valve depicting point of obstruction.\n\nIn retrospect, an LAO view during the coronary angiography would have been helpful to visualize the ostium of the LCA. At the time that the RAO view was performed, filling of the coronary arteries was satisfactory and the ostium itself was not clearly visualized. This angiography most likely demonstrated adequate flow through a partial obstruction. After hemodynamic deterioration, a subsequent angiography in the LAO view did visualize the ostium of the LCA (Figure 2C and Figure 4). The Medtronic CoreValve was then removed and the left coronary ostium was noted to be completely patent, leading to no further LCA intervention. As a result of this case, the routine use of LAO coronary angiography has been instituted in our facility.\n\nSeveral anatomic risk factors for coronary occlusion have been described in patients undergoing TAVR procedures. These include short distances to either coronary ostium (<10mm), shallow coronary sinuses, and narrow roots4,6. Our patient’s anatomy was appropriate for the placement of a Medtronic CoreValve; however, the distance to the LCA was just above 10 mm. This made the possibility of an occlusion after Medtronic CoreValve migration more likely.\n\n\nConclusion\n\nHere we report a case of a coronary artery osteal occlusion from the skirt of the core valve that converted into a total occlusion possibly from micro thrombi after protamine administration. This case highlights the vital role of echocardiography in differentiating between a protamine reaction and a coronary occlusion. It also emphasizes use of proper views during coronary angiography to diagnose coronary ostium occlusion, in the context of TAVR. We recommend vigilance and detailed reexamination of any hemodynamic instability after TAVR in order to expeditiously search for and identify impaired coronary blood flow, as well as any additional causes.\n\n\nConsent\n\nWritten informed consent was obtained from the patient for publication of this case report and any accompanying images and/or other details that could potentially reveal the patient’s identity.\n\n\nData availability\n\nFigshare: TEE short axis transgastric video. doi: 10.6084/m9.figshare.3507419.v19\n\nFigshare: TEE aortic valve long axis color video. doi: 10.6084/m9.figshare.3507428.v110\n\nFigshare: TEE four chamber video. doi: 10.6084/m9.figshare.3507431.v111\n\nFigshare: TEE Post Deployment Aortic valve long axis color. doi: 10.6084/m9.figshare.357120612",
"appendix": "Author contributions\n\n\n\nSB, TP, and KT conducted a literature search and prepared the manuscript. AO and KT prepared the multimedia files for the manuscript. JC and BG offered technical guidance for the manuscript. All authors were involved in the revision and drafting of the manuscript, and have agreed to the final format.\n\n\nCompeting interests\n\n\n\nThe following authors deny any potential conflicts of interest, including commercial relationships such as consultation and equity interests with any of the equipment mentioned in the article: Sujatha P. Bhandary, M.D., Andrew J. Otey, B.S., Thomas J Papadimos M.D.,MPH, Juan A. Crestanello, M.D., Katja R. Turner, M.D.\n\nThe author Barry S. George, M.D., would like to disclose the following financial relationships:\n\nCompensation and Benefits\n\n\n\n• (SFI) Medtronic, Inc.: $100–$999 (Travel, Lodging, and/or Meals)\n\n• (SFI) Boston Scientific: $100–$999 (Consulting/Advising/Employment)\n\nThe author verifies that these relationships did not affect or influence the preparation of this case report.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHamm CW, Möllmann H, Holzhey D, et al.: The German Aortic Valve Registry (GARY): in-hospital outcome. Eur Heart J. 2014; 35(24): 1588–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRibeiro HB, Nombela-Franco L, Urena M, et al.: Coronary obstruction following transcatheter aortic valve implantation: a systematic review. JACC Cardiovasc Interv. 2013; 6(5): 452–61. PubMed Abstract | Publisher Full Text\n\nSaia F, Marrozzini C, Marzocchi A: Displacement of calcium nodules of the native valve as a possible cause of left main occlusion following transcatheter aortic valve implantation. J Invasive Cardiol. 2011; 23(5): E106–9. PubMed Abstract\n\nMöllmann H, Kim WK, Kempfert J, et al.: Complications of transcatheter aortic valve implantation (TAVI): how to avoid and treat them. Heart. 2015; 101(11): 900–8, pii:heartjnl-2013-304708. PubMed Abstract | Publisher Full Text\n\nKukucka M, Pasic M, Dreysse S, et al.: Delayed subtotal coronary obstruction after transapical aortic valve implantation. Interact Cardiovasc Thorac Surg. 2011; 12(1): 57–60. PubMed Abstract | Publisher Full Text\n\nRibeiro HB, Sarmento-Leite R, Siqueira DA, et al.: Coronary obstruction following transcatheter aortic valve implantation. Arq Bras Cardiol. 2014; 102(1): 93–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDvir D, Leipsic J, Blanke P, et al.: Coronary obstruction in transcatheter aortic valve-in-valve implantation: preprocedural evaluation, device selection, protection, and treatment. Circ Cardiovasc Interv. 2015; 8(1): pii: e002079. PubMed Abstract | Publisher Full Text\n\nSingh M: Bleeding avoidance strategies during percutaneous coronary interventions. J Am Coll Cardiol. 2015; 65(20): 2225–2238. PubMed Abstract | Publisher Full Text\n\nBhandary S, Otey A, Papadimos T, et al.: TEE short axis transgastric video. Figshare. 2016. Data Source\n\nBhandary S, Otey A, Papadimos T, et al.: TEE aortic valve long axis color video. Figshare. 2016. Data Source\n\nBhandary S, Otey A, Papadimos T, et al.: TEE four chamber video. Figshare. 2016. Data Source\n\nBhandary S, Otey A, Papadimos T, et al.: TEE Post Deployment Aortic valve long axis color. Figshare. 2016. Data Source"
}
|
[
{
"id": "16581",
"date": "04 Oct 2016",
"name": "Wilbert Aronow",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting single case report that merits publication. I concur with the authors' approach to this patient.",
"responses": []
},
{
"id": "17789",
"date": "19 Dec 2016",
"name": "Kunal Sarkar",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a case of Left Coronary Artery shutdown post Core Valve Deployment that was attended with a cephalad migration of the prosthesis post deployment. While immediate post deployment angiogram did not reveal any limitation to coronary perfusion, the patient developed cardiogenic shock and global hypokinesis consistent with left coronary shutdown necessitating surgical bail out. I commend the authors for sharing their experience with this interesting case. I do have some critiques to offer regarding the report. I would suggest addressing these prior to indexing of this report.\nCore valve sizing is perimeter based. The authors should provide the amount of oversizing for 26mm Core valve based on manufacturer guidelines in the initial part of their report. The Left Coronary artery height was borderline although sinus of valsalva width was adequate.\n\nCephalad migration of Core Valve is not always a benign event. The device has three portions. An inflow portion that has the highest radial force, a constrained middle portion designed to allow unhindered coronary perfusion and an outflow portion for maintaining orientation with respect to the ascending aorta.Since Core valve is a nitinol based device it keeps expanding for 10-20 minutes post deployment. I would refer the authors to the findings of the core valve pivotal trial wherein the amount of paravalvar regurgitation was progressively reduced compared with immediate post deployment assessment. This is a function of continuous outward expansion of Core Valve and helps us avoid routine post dilation with BAV balloons. The timing of Coronary occlusion is consistent with a functional occlusion due to continuous expansion of the inflow portion. The presence of a skirt in the inflow portion is also a factor but the lack of thrombus at the time of surgical exploration argues more in favor of a compression of native aortic valvar tissue as the primary mechanism.\n\nAs the authors point out it is important to perform LAO (Caudal preferably) angiogram post deployment to assess left coronary ostium post deployment. It is not uncommon to have some mobilisation of aortic leaflet tissue in the region of the Left main ostium, however in most cases flow is preserved and \"functional\" left main stenosis is inconsequential.\n\nProtamine reaction is not a likely cause given the imaging and angiographic findings.\n\nThis report again underscores the need for meticulous preparation and anticipation prior to deployment and an understanding of device behavior to expeditiously solve post deployment issues.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2032
|
https://f1000research.com/articles/5-2000/v1
|
16 Aug 16
|
{
"type": "Opinion Article",
"title": "Top 10 metrics for life science software good practices",
"authors": [
"Haydee Artaza",
"Neil Chue Hong",
"Manuel Corpas",
"Angel Corpuz",
"Rob W.W. Hooft",
"Rafael C. Jiménez",
"Brane Leskošek",
"Brett G. Olivier",
"Jan Stourac",
"Radka Svobodová Vařeková",
"Thomas Van Parys",
"Daniel Vaughan",
"Haydee Artaza",
"Neil Chue Hong",
"Angel Corpuz",
"Rob W.W. Hooft",
"Rafael C. Jiménez",
"Brane Leskošek",
"Brett G. Olivier",
"Jan Stourac",
"Thomas Van Parys",
"Daniel Vaughan"
],
"abstract": "Metrics for assessing adoption of good development practices are a useful way to ensure that software is sustainable, reusable and functional. Sustainability means that the software used today will be available - and continue to be improved and supported - in the future. We report here an initial set of metrics that measure good practices in software development. This initiative differs from previously developed efforts in being a community-driven grassroots approach where experts from different organisations propose good software practices that have reasonable potential to be adopted by the communities they represent. We not only focus our efforts on understanding and prioritising good practices, we assess their feasibility for implementation and publish them here.",
"keywords": [
"Software",
"Best Practices",
"Metrics",
"Evaluation",
"Impact",
"ELIXIR",
"Sustainability"
],
"content": "Introduction\n\nCompliance with and promotion of good development practice is a powerful mechanism for promoting software sustainability. Using metrics to judge good practice can enhance research software maintainability and helps establish a baseline of quality, reusability and reproducibility. Software development metrics, however, are only useful if it is clear what they measure. This could be a) the application of agreed good practice in a piece of software or software team, or b) how sustainable the software is in the long term. There have been previous attempts to assess good practices for scientific computing1 but they did not specifically tackle the question how to measure them during software development. As part of a collaboration between the ELIXIR pan-European research infrastructure for life science data and the Software Sustainability Institute, a working group met at Schiphol airport (Amsterdam) on December 14–15th 2015 to a) define and select the metrics that reflect the application of good practices, b) discuss the collection of these metrics and c) establish how the metrics could be implemented to ensure their wide adoption. In this article we report the outcomes of this workshop. We believe this effort is set apart from previous initiatives because of its ‘bottom up’ approach to ensure community adoption and therefore it should have realistic chances of implementation. We benefit from the fact that participating members of both groups have long established track records in life science research software development. This is the first release of our agreed software development good practices and expect that new revisions could evolve from it in future iterations. It is outside of the scope of this manuscript to delve into the issues that these metrics might raise in terms of performance comparison between different software.\n\n\nMethods\n\nIn a workshop 12 experts from across Europe met to discuss good software development practices for life sciences. At the meeting, the group was divided randomly into two equally large subgroups to facilitate discussion, each subgroup spending a set time discussing potential metrics, their impact and applicability. The experts in each subgroup did not impose any restriction on which metrics to propose, but rather aimed to be as inclusive as possible, as long as each suggested metric had potential for impact. After the discussion, each group summarised the results and subsequently we merged the resulting metrics together into a list of 17 topics.\n\nNext, the two groups worked on prioritising the identified metrics according to two criteria: 1) Importance and 2) Implementability. Importance is a measure of the impact that a particular practice can have in making software more sustainable. A metric is considered highly implementable if it is easy to generate. For each identified metric, importance and implementability were ranked by all members of the working group on a scale from 1 to 5, 5 being highest importance or easiest implementation. An average score was calculated and the resulting list was sorted from highest to lowest scoring metrics. Here we discuss and evaluate a final list of the top ten prioritised metrics.\n\n\nResults\n\nWe identified a set of 17 topics that are critical for software development good practice (Box 1). It was evident that these include measurements of different styles: measurements can be self-reported, automatically produced or externally audited. The type of metric is also important to consider here: there are metrics of qualitative and quantitative nature. Qualitative metrics correspond to a binary classification description, while quantitative metrics tend to be more amenable to integration and presentation as statistics. Metrics interpretation may pose challenges of its own kind, particularly related to the subjective nature of the importance of metrics and the different perceptions of value according to the context in which they are used.\n\nEach of these topics have quantitative and qualitative metrics that may help track the adoption of good practice and monitor compliance with the guidelines in life sciences.\n\n1. Version control:\n\na. Yes/no?\n\nb. How many committers?\n\nc. When was the version control started?\n\nd. When was the last commit?\n\n2. Code reviews:\n\na. Yes/no?\n\nb. Star rating based on code description\n\n3. Automated testing:\n\na. Yes/no?\n\nb. Coverage for unit tests\n\nc. Yes/no for individual tests:\n\ni. Unit tests\n\nii. Functional tests\n\niii. Integration tests\n\niv. Regression tests\n\nd. Are the tests part of the code in the repository?\n\n4. Not reinventing the wheel:\n\na. Using libraries?\n\nb. Using Frameworks?\n\nc. Describing the algorithm, explaining why known code is reimplemented.\n\nd. Reinventing should be documented. References to the algorithm?\n\ne. Percentage of code written from scratch?\n\nf. Percentage of code that is involved in the core functionality?\n\n5. Discoverability:\n\na. Via structured search on functionality?\n\nb. Is it in the ELIXIR Tools and Data Services Registry2 or others (e.g., BioSharing3)?\n\n6. Reusability of source code:\n\na. Number or reuses = number of derived projects/external commits?\n\n7. Reusability of software:\n\na. Number of citations on the paper\n\nb. Having basic description of features in structured ELIXIR format (EDAM ontology4) - in ELIXIR Tools and Data Services Registry?\n\n8. Licensing:\n\na. Is there a license?\n\nb. Is the source available?\n\nc. Is it open source according to opensource.org?\n\n9. Issue tracking/bug tracking:\n\na. Does it have a publicly accessible issue tracker?\n\nb. How long are issues open?\n\nc. What is the number of unresolved issues?\n\nd. How much activity has there been in the last three months in the issue tracker?\n\n10. Support processes:\n\na. Are basic processes defined? Like governance, mailing list, releases, ...\n\n11. Compliance with community standards:\n\na. Yes/no?\n\nb. Specifies the level of compliance, specification version or metrics?\n\n12. Buildable code:\n\na. Does the compiler give warnings?\n\nb. Does a static analysis (“lint”) give warnings?\n\nc. Is an automated build system used?\n\n13. Open development:\n\na. Number of external committers in the repositories\n\n14. Making data available:\n\na. Yes/no?\n\nb. Where?\n\n15. Documentation:\n\na. Ratio code/comments, code lines/document lines?\n\nb. Percentage of code dedicated to documentation?\n\n16. Simplicity:\n\na. Measure of cyclomatic complexity\n\n17. Dependency management:\n\na. Is it done automatically using a system?\n\nb. Does it use a language-standard repository to pull in dependencies?\n\nc. Is software made available as a dependency in a dependency repository?\n\nWe used the 43 metrics contained in the 17 identified topics as a basis for further prioritisation as described in the Methods section. Prioritisation of metrics was achieved by all participants scoring them according to their perception of importance and implementability. An average score was calculated and a sum of average importance and average implementability to rank the list (Table 1). We introduced also a manual evaluation for each of the proposed ranked metrics, which reflected the consensus of the final prioritisation, given initial difference of opinions when reviewing the average scores. In Table 1, we summarise the top 10 suggested metrics.\n\nEach identified metric was scored according to importance (for sustainability) and implementability. Importance scores ranged from 1 (little) to 5 (very much) and implementability from 1 (difficult) to 5 (easy). Average values are shown for both importance (a) and implementability (b). A priority score (c) is calculated as the sum of the averages provided by (a) and (b). (c) is further discussed and the final Manual Priority Evaluation (d) is agreed, reflecting the final prioritisation judgement decided by the Working Group.\n\nAs a use case, we base the application of these metrics within the context of code development in ELIXIR. We define each of the 10 prioritised metrics in Table 1 and, where necessary, describe and explain the motivation for a metric and how to measure it. We consider that these definitions are applicable to a wider range of software development communities in life sciences.\n\n1. Is controlled versioning used?\n\n○ Description: Is it clearly indicated, can it be easily found?\n\n○ Motivation: Version control systems provide an environment for safe and transparent software development.\n\n○ How to measure: Put information about a version control tool to the ELIXIR Tools and Data Services Registry (which system, when it was installed, …)\n\n2. Is the software discoverable?\n\n○ Description: Is it easy to find the software based on its functionality (without knowing its exact name)?\n\n○ Motivation: It is important to be discoverable so other potential contributions are encouraged and more people use the software.\n\n○ How to measure: The ELIXIR community should be motivated and guided to provide this information into the ELIXIR Tools and Data Services Registry. If not, a list of other catalogues should be defined (maximum 5–10 other sources, e.g. BioSharing, field-specific catalogues, etc.). If the tool cannot be found there, the discoverability should be evaluated as 0.\n\n3. Is an automated build system used?\n\n○ Description: Are the builds of the software performed by some automated system?\n\n○ Motivation: If the automated system for builds is applied, can the users rebuild the software easily, which markedly increases its usability?\n\n○ How to measure: This information should be again included into the ELIXIR Tools and Data Services Registry2. Ideally, a link to the installation document should also be provided. How many commands are necessary for building of the software? (Optimally, just one command should be performed.)\n\n4. Are test data available?\n\n○ Description: Are data for testing of the software easily available for users?\n\n○ Motivation: Without test data, it may be difficult to try the functionality of the software and assess correct functioning of an installation.\n\n○ How to measure: The test data should be linked to from the web page describing the software or in the supplementary material of its associated publication. A link to the data should be included in ELIXIR Service Registry.\n\n5. Does software contain parts that reimplement existing technology?\n\n○ Description: Are common components/algorithms covered by libraries or reimplemented?\n\n○ Motivation: A (naïve) reimplementation can cause unnecessary errors or decrease the effectiveness.\n\n○ How to measure: Percentage of code written from scratch and/or number of used libraries. Additionally, descriptions of why a library with similar functionality was not used and responses to suggestions from community.\n\n6. Does the software support open community standards and what is its level of compliance?\n\n○ Description: Evaluation of software compliance with open/community standards\n\n○ Motivation: This is needed, for example, where data input/output, networking and general interoperability are concerned. However, it is also non-trivial to implement and measure in terms of the overall software quality.\n\n○ How to measure: A base metric would be: “does the software make use of open standards (yes/no), if so which ones (listing)?” In addition, more qualitative information such as “which versions of the standard does the software support?”, “Is it compatible with the latest specification?”, and “Can it be used to provide a more general level of support?” Another fundamental aspect to consider is whether the standard provide its own compliance metric (e.g., a test suite) and what the software’s level of compliance is. An example of such a compliance test suite is provided by the Systems Biology Markup Language (SBML,5).\n\n7. Are code reviews performed?\n\n○ Description: Whether new code is inspected by someone else before it becomes part of the code base.\n\n○ Motivation: Code reviews increase quality of the code both because it is written with more care and because the second pair of eyes will more readily catch false assumptions or errors.\n\n○ How to measure: Activity in code review process (comments to updated lines, etc.)\n\n8. Is automated testing performed?\n\n○ Description: Is some system for automated testing implemented?\n\n○ Motivation: Automatic testing decreases occurrence of bugs.\n\n○ How to measure: Information about the testing methodology should be present in the software documentation. In parallel, developers can be motivated to add this information to ELIXIR Service Registry.\n\n9. Is the code documented?\n\n○ Description: Does the code contain comments describing its main elements?\n\n○ Motivation: Code comments increase the readability of the code and also indirectly motivate the programmer to write a cleaner code. However, commenting can present the problem of not being updated as code changes. This means that code comments may rot and become misleading/inaccurate. Often comments can be made redundant by better names of variables and methods. An exception is example code where explaining what each line does with a comment is useful.\n\n○ How to measure: Determine the percentage of text from the source code that corresponds to comments.\n\n10. How high is the code complexity?\n\n○ Description: This refers to how complex or straightforward the code is.\n\n○ Motivation: The more complex code, the higher risk of errors. Code can be simplified by proper separation of tasks into different routines and methods.\n\n○ How to measure: Measure the cyclomatic complexity.\n\n\nDiscussion and conclusion\n\nWe present an initial set of 10 good practices that could help make software for the life sciences more sustainable. From our discussions, it was clear that a community-wide adoption of standards is needed in terms of how measurement of metrics are collected and shared. We operate under the assumption that all software developed should be open source from the beginning of development, which means that the collection of statistics for good practice compliance should not violate any of the licensing or privacy issues associated to closed code.\n\nThese ‘Top 10 Good Practices’ should be considered as an initial view of what the community considers important with a description of their feasibility for implementation within the life sciences. Among our top suggested topics there is a remarkable coincidence on the need for versioning. The ways on how to collect metrics regarding versioning systems vary: if using GitHub, a number of statistics are readily available that allow their easy collection for benchmarking. We do not, however, want to prescribe which versioning systems should be adopted. There are many ways in which this metric can be measured, a sample of which we offer. The metrics we propose can be both qualitative and quantitative. Although quantitative metrics are easier to track, it is also important to capture qualitative characteristics such as existence of automated testing or compliance with community standards.\n\nThis article is a first attempt to crystallise the conclusions from the work that the group of experts gathered under the auspices of ELIXIR and the Software Sustainability Institute. It is thus not intended to be a final declaration of what the ELIXIR community thinks the metrics, implementation and feasibility for measuring good practices for software development should be. This document is an initial response from the working group established to assess the problem of evaluating metrics for software development good practices. We expect it to be modified in future versions as more experts join this group and new challenges emerge with evolving technologies and life science software needs.",
"appendix": "Author contributions\n\n\n\nAll authors participated in the discussions, selection and prioritisation of metrics. We believe all authors contributed equally to this work. All authors contributed to the writing of this article. All authors read and approved the submitted manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nMC and HA are strategically core funded by UK’s BBSRC. BGO is funded by the BE-BASIC grant F08.005. NCH was supported by EPSRC, BBSRC and ESRC Grant EP/N006410/1 for the UK Software Sustainability Institute. The work was part of the ELIXIR-EXCELERATE project, funded by the European Commission within the Research Infrastructures programme of Horizon 2020, grant agreement number 676559.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe are grateful to Susanna Repo (ELIXIR-Hub) and Chris Ponting (ELIXIR-UK Head of Node) for comments and feedback.\n\n\nReferences\n\nWilson G, Aruliah DA, Brown CT, et al.: Best practices for scientific computing. PLoS Biol. 2014; 12(1): e1001745. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIson J, Rapacki K, Ménager H, et al.: Tools and data services registry: a community effort to document bioinformatics resources. Nucleic Acids Res. 2016; 44(D1): D38–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nField D, Sansone S, Delong EF, et al.: Meeting Report: BioSharing at ISMB 2010. Stand Genomic Sci. 2010; 3(3): 254–258. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIson J, Kalas M, Jonassen I, et al.: EDAM: an ontology of bioinformatics operations, types of data and identifiers, topics and formats. Bioinformatics. 2013; 29(10): 1325–1332. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHucka M, Bergmann FT, Dräger A, et al.: Systems Biology Markup Language (SBML) Level 2 Version 5: Structures and Facilities for Model Definitions. J Integr Bioinform. 2015; 12(2): 271. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15707",
"date": "24 Aug 2016",
"name": "Pedro L. Fernandes",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a pragmatic report that describes a systematic approach to selecting criteria to measure the adoption of good practices in software development. The context of this study is a not as wide as the title and abstract indicate. However, the assessments are fair and the methodology for collecting data is well described. The use case provides a ground for a concrete reflection, and that is very useful. The generalisation of the method will require more decoupling from the use case, but it is understood that it can happen in future iterations of this study.\nIt is reasonable to believe that various grades of agreement has been reached in the discussions. It would have been interesting to have included some measurement of dispersion, as the prioritisation results from ranking average scores alone (Table 1). It would add value to the confidence in the choice of the top 10 prioritised metrics that otherwise look indistinguishable from a full consensus. Although it is useful to concentrate on the top 10 and describe them at this level of detail, it would be very useful to see the ranking of the full set of 43 metrics analysed, possibly in a bar graph with the average score and error bars to indicate dispersion. That would enrich this manuscript significantly.\nThe importance of this effort in defining such criteria is very large, and, as the authors suggest, it represents a first step of an iterative process that is much needed in this area. In subsequent iterations some refinement of the measurement methods will be needed, such as in metric #9. \"Is the code documented?\", presently listed as \"How to measure: Determine the percentage of text from the source code that corresponds to comments.\" The advantage of having the metric in usage is high, but the way to measure needs to match the relative importance of the criteria. In this case, a high percentage of text in comments is relevant is the comments are useful and contextualised, and it can be argued that a large quantity of irrelevant comments can actually be detrimental in various aspects, from code readability to maintainability itself.",
"responses": []
},
{
"id": "15704",
"date": "26 Aug 2016",
"name": "Bruno Gaëta",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAs a meeting report and opinion piece, there is little to contradict in the manuscript. The ideas put forward are sensible and make a great starting point for discussion which is the purpose of the manuscript. Maybe the scope of the article could be slightly better defined to differentiate the goal of the proposed metrics from that of the metrics and methods used for software verification (see for example Giannoulatou et al (2014) )",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-2000
|
https://f1000research.com/articles/5-1989/v1
|
15 Aug 16
|
{
"type": "Research Article",
"title": "Quality of surgical care of pancreatic cancer in a single payer North American health care system",
"authors": [
"Scott Hurton",
"Robin Urquhart",
"Cynthia Kendall",
"Margaret Jorgensen",
"Geoff Porter",
"Adrian Levy",
"Michele Molinari",
"Scott Hurton",
"Robin Urquhart",
"Cynthia Kendall",
"Margaret Jorgensen",
"Geoff Porter",
"Adrian Levy"
],
"abstract": "Introduction: Quality of surgical care of Canadian patients with pancreatic cancer (PC) is inadequately known. Primary aim of this study was to analyze the quality of care received by patients who underwent resections of PC in Nova Scotia over a 10-year period (2001-2011). Methods: All patients with PC (n. 1094) were identified using provincial cancer registries and only adult patients with resectable disease were included in the study (n. 109). Well established disease-specific quality indicators (QIs) were used as references. The proportion of patients who met those QIs was calculated. The average and 95 % confidence intervals of QIs were compared between patients treated in Nova Scotia and published references. Results: Surgical therapy was performed in 9.9 % of patients. Perioperative morbidity and mortality occurred in 25 % and 5 % of patients respectively. Overall survival was 57 % at 1 year, 18 % at 3 year and 9 % at 5 year. R1 resections occurred in 38 % of patients. When compared to published quality measures, patients in Nova Scotia had similar outcomes except for an inferior number of lymph nodes found in the surgical specimens (9 vs. 19; P<0.05). On the other hand, a significant proportion of patients did not fully meet several QIs linked to preoperative, surgical pathology and postoperative care. Conclusions: In Nova Scotia, the proportion of patients who underwent surgery for PC was lower than expected. Although perioperative morbidity, mortality and survival rates were comparable to published references, many did not meet established standard QIs.",
"keywords": [
"pancreatic cancer",
"surgical care",
"Nova Scotia",
"quality indicators",
"mortality",
"survival"
],
"content": "Introduction\n\nPancreatic cancer (PC) is one of the most common gastrointestinal malignancies in North America and Europe1,2. In comparison to many other malignancies, it has poor prognosis due to late diagnosis and low response to current chemo-radiation therapies2. In addition to aggressive biology, recent data also suggest that, within different health systems, the quality of care provided to PC patients is heterogeneous and often suboptimal3–11. In Canada, national data published in 201412 revealed that the overall 5-year survival of patients diagnosed with PC correlated with the geographical areas of their residence, with Nova Scotia having the lowest survival rate (4.7% 5-year survival in comparison to the national rate of 9.1%; P<0.05)13. Several hypotheses have been raised to explain these findings11: lower socio-economic conditions, higher alcohol and tobacco consumption, more sedentary lifestyle and higher prevalence of obesity than in other Canadian provinces14–16. Alternatively, these results might be due to lower quality of care delivered to PC patients living in Nova Scotia where, contrary to several other Canadian provinces, a formal and coordinated quality improvement intervention for PC had never occurred11.\n\nThe primary aim of this study was to compare the quality of care and outcomes of patients who underwent surgical therapy for PC in Nova Scotia over a 10-year period using disease-specific quality indicators (QIs) published in the scientific literature as references11,17.\n\n\nMethods\n\nThe study population included a cohort of patients older than 18 years of age who had undergone resection for primary exocrine pancreatic cancers (PC) in the province of Nova Scotia, Canada over a 10-year period (April 1, 2001–March 31, 2011). Diagnostic codes of the International Classification of Disease for Oncology, 3rd edition (ICD-O-3) were used to select patients with PC from the Nova Scotia Cancer Registry (NSCR), a prospectively maintained provincial registry of all patients affected by malignant diseases in the province. Diagnostic codes of malignancies of the exocrine pancreas used for this study are summarized in Table 1. Patients were excluded if affected by pancreatic endocrine neoplasms, lymphomas, sarcomas, metastases from other malignancies or direct invasion of tumors originating in surrounding organs (e.g. retroperitoneum, gastrointestinal tract).\n\nEach patient who satisfied the inclusion criteria was assigned an identification (ID) number to protect patients’ privacy. Additional administrative datasets were linked to the NSCR for completion of sociodemographic and clinical data as represented in Figure 1. Linked datasets included the Oncology Patient Information System (OPIS), Medical Service Insurance (MSI) Physician Services, Medical Service Insurance (MSI) Patient Registry, and the Canadian Institutes of Health Information (CIHI) Discharge Abstract Database (DAD)18. These datasets were linked to the NSCR by the Population Health Research Unit (PHRU) at Dalhousie University. The methodology used to link these administrative datasets has been described in details in previous manuscripts published by our group17,18. Surgical patients were excluded if they underwent palliative interventions such as biliary or gastric bypasses when found to be unresectable intraoperatively.\n\nLegend: NSCR (Nova Scotia Cancer Registry), HCN (Health Care Number), CHRD (Capital Health Radiology Department), PHRU (Population Health Research Unit, Dalhouse University).\n\nApproval for this study identified as CDHA-RS/2012-206 was obtained from Capital Health Research Ethics Board Centre for Clinical Research located in room 322B, CCR at 5790 University Avenue, Halifax, NS B3H 1V7, Canada. This ethic review board supervises the good conduct of research projects performed by investigators with appointments at Dalhousie University. In addition, the protocol was also approved by all the ethic review boards (ERB) at each provincial health district where patients received their treatment. Besides Capital Health Research Ethics Board, other ERBs responsible for the approval of this study were: Annapolis Valley, South Shore and South West Nova Scotia, Cape Breton and Guysborough Antigonish Strait, Colchester-East Hants, Cumberland and Pictou Health Authorities19,20.\n\nThe quality of surgical care delivered to patients with PC was measured using QIs proposed by Sabater et al.17 and Bilimoria and colleagues11. Quality measures selected from Sabater’s study21 were: perioperative morbidity, perioperative mortality, overall 1, 3 and 5-year survival, number of lymph nodes reported within the surgical specimen and state of resection margins. The international classification of surgical pathologists was used to describe the involvement of resection margins with R0 margins meaning absence of cancer cells seen microscopically, R1 indicating that cancerous cells could be seen microscopically and R2 when tumor tissue was visible at naked eye on the margins at gross examination. Overall survival was recorded at 1, 3, and 5 years after surgery. All causes of death were considered secondary to recurrent disease and patients who were still alive at the closure of this study were censored. The only variation that occurred in reporting perioperative complications in this study was that adverse events were recorded only if satisfied grade III to V of Clavien-Dindo classification21 contrary to Sabater et al.17 who recommended reporting all grades of perioperative complications. QIs identified from the study by Bilimoria et al.11 were summarized in four domains: preoperative, operative, surgical pathology reporting and postoperative outcomes.\n\nThe quality of care of patients who underwent surgical resection of PC in Nova Scotia was compared to the mean and 95% confidence interval (CI) of QIs selected by Sabater et al.17. In their study, Sabater and colleagues17 performed a comprehensive search of practice guidelines, consensus conferences and reviews of pancreatic oncologic surgery and selected clinical relevant indicators of quality with weighted averages and respective 95% CIs. Comparisons between Nova Scotia and outcome benchmarks were performed using 95% CI for each of the QIs to test for possible statistically significant differences.\n\nOverall survival analysis was performed using Kaplan-Meier methodology22. All statistical analysis was carried out using SAS® (Version 8.2, Cary, North Carolina). Two-tailed analyses were performed unless otherwise specified. Missing data were excluded except when imputation was possible from administrative data. All statistical analyses were considered significant at p<0.05.\n\n\nResults\n\nInclusion criteria were satisfied by 109 patients. Fifteen patients (13.7%) were excluded due to significant missing data. As a result, a total of 94 patients were included and represented the study population where the median age was 66.8 years, 41% were female, and 87% underwent a pancreaticoduodenectomy (Table 2).\n\nPerioperative complications occurred in 25 patients (24.5%) (95% CI: 16.2–32.8%), and perioperative mortality occurred in 5 (5.9%) (95% CI: 2.2–10.5%). R1 resections occurred in 38 patients (37.3%) (95% CI: 37.2–46.6%) (Figure 2). The mean number of lymph node retrieved in each specimen was 9.0 (95% CI: 7.7–10.3) and the overall patient survival at 1, 3, and 5 years was 55.1% (95% CI: 45.2–63.8%), 18.5% (95% CI: 11.5–26.7%), and 9.4% (95% CI: 4.2–17.1%), respectively (Figure 2). When data from Nova Scotia were compared to values reported by Sabater et al.17, there were no statistically significant differences as 95% CIs overlapped between the two groups except for the mean number of lymph nodes identified in the pathology specimens and the rate of serious complications that were lower for patients operated in Nova Scotia.\n\nThe black squares and horizontal lines represent the references used to compare the outcome of patients who underwent surgical care in Nova Scotia (Red squares and lines). Statistical significant differences are present only if the horizontal lines representing 95% confidence intervals between the two groups do not overlap.\n\nAssessment of other QIs proposed by Bilimoria et al.11 revealed that there was significant heterogeneity in the percentage of Nova Scotia patients who met the QIs across all four different domains (Table 3). For example, in the preoperative domain, 82% of patients had an appropriate cross-sectional imaging study within 2 months from the day of their surgery and 93% underwent treatment within 2 months after their diagnosis. On the other hand, only 33% of patients had 10 or more lymph nodes identified in the surgical specimen and only 24% had a complete TNM stage description in the final surgical pathology report.\n\nQIswere grouped in four domains: preoperative, operative, surgical pathology and postoperative care. The number and percentage of patients who underwent pancreatic resections in Nova Scotia and who met the QIs proposed by Bilimoria et al.11 are reported in the right column.\n\n\nDiscussion\n\nIn 2014, Statistic Canada12 published that in Nova Scotia, 5-year overall survival of PC patients was 4.7% in comparison to the national rate of 9.1%23. These results confirmed observations from other researchers who reported geographical variations in how PC patients are treated and differences in their overall outcomes. Variations treatments depend on multiple factors including physicians’ expertise, hospital resources, and patients’ characteristics. For example, other researchers have shown that patients who live in rural areas or who belong to lower socio-economic groups have worse outcomes when diagnosed with complex gastrointestinal malignancies or other chronic diseases13,14,24–26. This might be due to barriers to access hospitals or specialists, especially in countries where for-profit organizations play an important role in the delivery of healthcare. In Canada, these potential barriers should not exist as health care services are public and, ideally, equally accessible to all citizens.\n\nIn a previous paper11, our group suggested the possibility that the lower overall survival of PC patients living in Nova Scotia was due to differences in their socio-demographic characteristics when compared to other Canadian provinces14–16. Unfortunately, the data provided by Statistic Canada were not sufficiently granular to assess patients’ characteristics, and we could not exclude that patients diagnosed with PC in Nova Scotia received suboptimal care. Because of these concerns, we carried out an extensive epidemiological study that included all patients diagnosed with PC over a 10-year period in Nova Scotia with the main intent of assessing the quality of their care using established disease-specific indicators.\n\nOne of the main findings of this study was that a small proportion of patients underwent surgical treatment during the study period. In fact, among a total number of 1094 patients diagnosed with PC over a ten-year period, only 109 (9.9%) underwent radical surgery. This is in contrast to reports from other Canadian centers where radical surgery was feasible in up to 25% of referrals15. Our study, however, has shown that the quality of care of patients who underwent surgery was within the confidence intervals of benchmarks published by Sabater et al.17. In other words, there were no significant statistical differences in perioperative mortality, R1 resection rates and overall survival at 1, 3 and at 5-year after surgery between our population and the parameters used for comparison. Nevertheless, there was a trend towards higher perioperative mortality, and overall lower survival rate compared to the pooled data from the scientific literature.\n\nWhen we analyzed the proportion of patients who met QIs in the four domains proposed by Bilimoria et al.11, we found that there were considerable gaps in the quality of surgical pathology reporting and utilization of adjuvant therapies. The quality of pathology reports might have influenced how patients were managed after their operation. In fact, patients with surgical pathology reports that did not mention positive resection margins or lymph node involvement were rarely referred to medical oncologists after their surgeries. The main reason was that in our institution, referral patterns, and medical oncological therapies were not unified and some providers perceived that there was no benefit for postoperative chemotherapy except for patients at high risk of recurrence (e.g. positive resection margins or positive lymph node involvement). Overall, only 53% of patients in our study were referred to medical oncology service for adjuvant therapy, and only 46% ended up to be seen by a medical oncologist.\n\nThis study has several limitations due to the retrospective nature of its design and the linking of multiple datasets that could have increased the risks of inaccuracy of the data as the primary purpose of these datasets was not for research16. Another important limitation is the small number of patients who were included. The overall number of inhabitants in Nova Scotia has been relatively stable with an estimated population of 930,000. Since the yearly incidence of PC in Canada is in the range of 8–11 patients/100,000 inhabitants18, we are confident that the majority of patients diagnosed during the study period were identified. However, the resectability rate was much lower than expected as the majority of referrals were unable to undergo surgery due to the advanced stages of the disease or the presence of severe comorbidities that precluded resections. Since surgery remains the only possible curative treatment for PC, the fact that less than 10% of patients were offered this opportunity raises the doubts that a considerable proportion was diagnosed late or not referred for surgical opinion in a timely fashion.\n\nDespite the above limitations, our study is original and, to the best of our knowledge, it is the first to assess the quality of care provided to patients who underwent surgery for PC at a Canadian provincial level by using established indicators. Also, before this study, there were no reports on the resectability rate of PC in Nova Scotia or the overall 5-year survival of patients who underwent surgery.\n\nIn conclusion, patients diagnosed with PC and treated with radical resections in Nova Scotia had acceptable outcomes comparable to published international standards. However, we found that resectability rate was below average and a significant proportion of patients did not meet several QIs including preoperative radiological studies that were older than two months, incomplete surgical pathological reporting and, finally, a low referral rate and utilization of adjuvant therapy. Because survival of patients with PC remains disappointing, it is important to monitor for correctable deficiencies in the health care system. Our study suggests that, in Nova Scotia, there is need to increase early detection and early surgical referral, a more exhaustive pathology reporting and an increased use of adjuvant chemotherapy. A coordinated implementation of all these interventions might improve the overall survival of PC patients in the province. However, areas in which quality improvement strategies are most effective remain unknown and should be explored in future studies.\n\n\nData availability\n\nNova Scotia's Personal Health Information Act, S.N.S. 2010, c 41 (PHIA), a provincial legislation which governs the collection, use, disclosure, retention and disposal and destruction of personal health information of patients treated in Nova Scotia, came into force on June 1 201319. The goal of this statute is to balance the privacy rights of individuals with respect to their personal health information and the need for researchers and practitioners to collect, use and disclose personal health information as part of providing strong healthcare services in Nova20. Due to restrictions imposed by PHIA, the raw data used for this study can only be accessible to investigators who obtain permission from Capital Heatlh Research Ethics Board Centre for Clinical Research and from the Nova Scotia Cancer Registry to use the encrypted database.\n\nInvestigators should submit their request to access the encrypted database used for this study to: Capital Health Research Ethics Board, Room 322, CCR, 5790 University Avenue, Halifax, NS, B3H 1V7, Canada.",
"appendix": "Author contributions\n\n\n\nScott Hurton performed the statistical analysis of the data, performed a systematic review of the literature, wrote the paper and created the figures and tables reported in the manuscript. In addition, he manually searched medical records when data were missing for some of the patients who were included in this study.\n\nRobin Urquhart, Cynthia Kendall, Margaret Jorgensen, Geoff Porter, Adrian Levy: Designed the protocol, assisted in linking all the different administrative provincial datasets used for this study and revised the manuscript.\n\nMichele Molinari: Wrote the study protocol, obtained grant funding to support this study, performed a systematic review of the literature, co-wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by a grant of $CAN 100,000 provided by Craig’s Cause Pancreatic Cancer Society, a registered Canadian Charity Society, Business Number: 842352759RR0001, webpage: http://www.craigscause.ca/.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to give a special thank to Stefanie Condon-Oldreive and Paul Oldreive (Founders of Craig’s Cause Pancreatic Cancer Society) for their generous support that provided all the necessary funds for this study. In addition, we would like to thank the Canadian Institutes of Health Research (CIHR) for the grant given to Dr. Scott Hurton for his Master Degree studies in Community Health and Evaluation at Dalhousie University.\n\nThe authors would like to thank all the volunteers of Craig’s Cause Pancreatic Cancer Society for their generosity and fund-raising activities, and Dr. Mark Walsh and Dr. Brock Vair for they operated on many patients who were included in this study.\n\n\nReferences\n\nEpidemiology of Pancreatic Cancer in the World. 2014. (Accessed 10 October, 2014). Reference Source\n\nSharma C, Eltawil KM, Renfrew PD, et al.: Advances in diagnosis, treatment and palliation of pancreatic carcinoma: 1990–2010. World J Gastroenterol. 2011; 17(7): 867–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim C, Owen D, Gill S: Real-world impact of availability of adjuvant therapy on outcomes in patients with resected pancreatic adenocarcinoma: a Canadian Cancer Agency experience. Am J Clin Oncol. 2012; 35(3): 212–5. PubMed Abstract | Publisher Full Text\n\nMerkow RP, Bilimoria KY, Bentrem DJ, et al.: National assessment of margin status as a quality indicator after pancreatic cancer surgery. Ann Surg Oncol. 2014; 21(4): 1067–74. PubMed Abstract | Publisher Full Text\n\nRiall TS, Eschbach KA, Townsend CM Jr, et al.: Trends and disparities in regionalization of pancreatic resection. J Gastrointest Surg. 2007; 11(10): 1242–51; discussion 51–2. PubMed Abstract | Publisher Full Text\n\nSimons JP, Shah SA, Ng SC, et al.: National complication rates after pancreatectomy: beyond mere mortality. J Gastrointest Surg. 2009; 13(10): 1798–805. PubMed Abstract | Publisher Full Text\n\nSimunovic M, To T, Theriault M, et al.: Relation between hospital surgical volume and outcome for pancreatic resection for neoplasm in a publicly funded health care system. CMAJ. 1999; 160(5): 643–8. PubMed Abstract | Free Full Text\n\nSimunovic M, Urbach D, Major D, et al.: Assessing the volume-outcome hypothesis and region-level quality improvement interventions: pancreas cancer surgery in two Canadian Provinces. Ann Surg Oncol. 2010; 17(10): 2537–44. PubMed Abstract | Publisher Full Text\n\nBilimoria KY, Bentrem DJ, Feinglass JM, et al.: Directing surgical quality improvement initiatives: comparison of perioperative mortality and long-term survival for cancer surgery. J Clin Oncol. 2008; 26(28): 4626–33. PubMed Abstract | Publisher Full Text\n\nBilimoria KY, Bentrem DJ, Tomlinson JS, et al.: Quality of pancreatic cancer care at Veterans Administration compared with non-Veterans Administration hospitals. Am J Surg. 2007; 194(5): 588–93. PubMed Abstract | Publisher Full Text\n\nBilimoria KY, Bentrem DJ, Lillemoe KD, et al.: Assessment of pancreatic cancer care in the United States based on formally developed quality indicators. J Natl Cancer Inst. 2009; 101(12): 848–59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCanadian Cancer Society's Advisory Committee on Cancer Statistics: Canadian Cancer Statistics 2014. Canadian Cancer Society, 2014; (Accessed May 2, 2014). Reference Source\n\nPancreas Cancer Mortaltity by Geography, Age Group and Gender. Canada.gc.ca, 2014; (Accessed March 26, 2014).\n\nBilimoria KY, Bentrem DJ, Ko CY, et al.: National failure to operate on early stage pancreatic cancer. Ann Surg. 2007; 246(2): 173–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOlivié D, Lepanto L, Billiard JS, et al.: Predicting resectability of pancreatic head cancer with multi-detector CT. Surgical and pathologic correlation. JOP. 2007; 8(6): 753–8. PubMed Abstract\n\nAyanian JZ: Using administrative data to assess health care outcomes. Eur Heart J. 1999; 20(23): 1689–91. PubMed Abstract | Publisher Full Text\n\nSabater L, García-Granero A, Escrig-Sos J, et al.: Outcome quality standards in pancreatic oncologic surgery. Ann Surg Oncol. 2014; 21(4): 1138–46. PubMed Abstract | Publisher Full Text\n\nCanadian Institute for Health Information: Discharge Abstract Database (DAD) Metadata. 2015; (Accessed September 9, 2011). Reference Source\n\nPersonal Health Information Act (“PHIA”). 2013; (Accessed 07, 2016). Reference Source\n\nCancer Care Nova Scotia. email: info@ccns.nshealth.ca. 2016. Reference Source\n\nDindo D, Demartines N, Clavien PA: Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004; 240(2): 205–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Amer Stat Assn. 1958; 53(282): 457–81. Publisher Full Text\n\nHurton S, MacDonald F, Porter G, et al.: The current state of pancreatic cancer in Canada: incidence, mortality, and surgical therapy. Pancreas. 2014; 43(6): 879–85. PubMed Abstract | Publisher Full Text\n\nEnewold L, Harlan LC, Tucker T, et al.: Pancreatic cancer in the USA: persistence of undertreatment and poor outcome. J Gastrointest Cancer. 2015; 46(1): 9–20. PubMed Abstract | Publisher Full Text\n\nSun H, Ma H, Hong G, et al.: Survival improvement in patients with pancreatic cancer by decade: a period analysis of the SEER database, 1981–2010. Sci Rep. 2014; 4: 6747. PubMed Abstract | Publisher Full Text\n\nMurphy MM, Simons JP, Ng SC, et al.: Racial differences in cancer specialist consultation, treatment, and outcomes for locoregional pancreatic adenocarcinoma. Ann Surg Oncol. 2009; 16(11): 2968–77. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15690",
"date": "05 Sep 2016",
"name": "Rachel Foskett-Tharby",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis was an interesting and well written article but the following points of data accuracy and interpretation should be addressed.\nTable 2 appears to have some missing data. Specifically the numbers of patients reported against each income quartile do not sum to 94. I'm also not sure if the numbers of patients experiencing a grade 5 perioperative complication have been missed from the table.\n\nThere is an inconsistency in the numbers of patients with perioperative complications reported in the text on page 5 (25 patients) and those reported in Table 2 (19 patients with grade 3 and 4 classifications of complications).\n\nThe final sentence of paragraph 2 of the results states that there was a statistically significant difference in perioperative morbidity between Nova Scotia and the Sabater et al study.1 Surely this has occurred not as a result of differences in practice but due to the differences in definition of perioperative morbidity between the studies. This comparison should be recalculated using the same definition (i.e. all grades of the Dindo-Clavien Classification of Perioperative Adverse Events) as the Sabater et al, study and appropriate conclusions drawn.",
"responses": [
{
"c_id": "2170",
"date": "07 Sep 2016",
"name": "Michele Molinari",
"role": "Author Response",
"response": "Dear Rachel Foskett-Tharby, Thank you for your review and for your comments. Comment 1- Table 2: there has been an error as 3 patients did not have information on their income quartile. The table is missing a row reporting the number of patients with missing value. We will resubmit the table with corrections. Comment 2- Table 2: there has been an error in reporting complications as the number of patients with grade 0-III Dindo Clavien complications should be 18 (19.1%) and with grade IV Dindo Clavien complications should be 6 (6.3%). Comment 3- The differences in perioperative complications between the population operated in Nova Scotia and the values reported by Sabater et al. are due to differences in definitions. In our study, grade I-II were not captured. Therefore, we were not able to include all those complications that did not have any significant clinical impact. Unfortunately, due to the retrospective design of our study, these minor complications were not available for statistical comparison."
}
]
},
{
"id": "16781",
"date": "04 Oct 2016",
"name": "Sulaiman Nanji",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study examines the quality of surgical care in patients undergoing surgical resection for pancreatic adenocarcinoma over a 10-year period (2001-2011) in the Canadian Province on Nova Scotia. The authors use the provincial cancer registry to identify the study cohort based on ICD codes. The quality of surgical care delivered to patients was assessed using previously reported, established disease-specific quality indicators including surgical factors, pathologic measures and short and long-term outcomes. The objective was to determine if patients undergoing curative-intent resection for pancreatic cancer in Nova Scotia were receiving similar quality of care compared to published references.\n\nOur surgical knowledge in the management and outcomes in pancreatic cancer is largely based on retrospective data from large volume centres, often single-institution based series. While these reports provide essential information, it is unclear whether similar outcomes are being realized in routine clinical practice. Health services research using population-based studies, such as this one, provide insight into the management and outcomes achieved in the “real world” and are less limited by referral and selection biases that are inherent in traditional institutional-based series. Moreover, they provide insight into access to care and the adoption of new treatments in routine practice, allowing for identification of gaps and deficiencies in the delivery of appropriate care.\n\nIn this study, the authors report inferior outcomes in patients in Nova Scotia compared to published quality measures. Specifically, the resectability rate was below average with only 10% of patients diagnosed with pancreatic cancer actually undergoing surgical resection with curative intent. They also found a below average rate of referral to medical oncology for consideration of adjuvant therapy, with only half of patients being referred. These are important findings as they identify critical deficiencies in the health care system that are correctable. It is conceivable that with improvements in referral to both surgical and medical oncology in Nova Scotia, long-term outcomes may improve, which at present are substantially worse compared to the reported rates in the literature (5-year overall survival 9% compared to 16%). The authors also identified critical gaps in the quality of surgical pathology reporting with poor reporting of margin status, as well as an inferior surgical node retrieval rate with only one-third of patients having > 10 lymph nodes resected. Again, these findings identify important areas of quality improvement.\n\nGiven the retrospective, administrative nature of the data, there are inherent limitations such as the lack of detailed information related to post-operative complications and factors that may have influenced referral to surgical and medical oncologists. Moreover, despite using registry data for the entire province of Nova Scotia over 10 years, the actual study cohort is quite small (n=94), limiting more detailed analysis. Perhaps there is a temporal relationship between quality of care and time, with better outcomes in the more recent era? Given the small numbers, it is not possible to explore how quality may have changed over time. Nevertheless, the authors should be commended for the candid reporting of this data, highlighting the insufficiencies in the delivery of care to patients with pancreatic cancer in Nova Scotia. Through their investigation they were able to identify several areas where quality of care can be improved, with a view to achieving better outcomes. This report also underscores the importance of health services research and population-based studies to determine if outcomes in routine practice are indeed comparable to published reports and to identify deficiencies in the health care system in order to better organize, manage and deliver quality care.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1989
|
https://f1000research.com/articles/5-1961/v1
|
12 Aug 16
|
{
"type": "Research Article",
"title": "Visual acuity measured with luminance-modulated and contrast-modulated noise letter stimuli in young adults and adults above 50 years old",
"authors": [
"Pui Juan Woi",
"Sharanjeet Kaur",
"Sarah J. Waugh",
"Mohd Izzuddin Hairol",
"Pui Juan Woi",
"Sharanjeet Kaur",
"Sarah J. Waugh"
],
"abstract": "The human visual system is sensitive in detecting objects that have different luminance level from their background, known as first-order or luminance-modulated (LM) stimuli. We are also able to detect objects that have the same mean luminance as their background, only differing in contrast (or other attributes). Such objects are known as second-order or contrast-modulated (CM), stimuli. CM stimuli are thought to be processed in higher visual areas compared to LM stimuli, and may be more susceptible to ageing. We compared visual acuities (VA) of five healthy older adults (54.0±1.83 years old) and five healthy younger adults (25.4±1.29 years old) with LM and CM letters under monocular and binocular viewing. For monocular viewing, age had no effect on VA [F(1, 8)= 2.50, p> 0.05]. However, there was a significant main effect of age on VA under binocular viewing [F(1, 8)= 5.67, p< 0.05]. Binocular VA with CM letters in younger adults was approximately two lines better than that in older adults. For LM, binocular summation ratios were similar for older (1.16±0.21) and younger (1.15±0.06) adults. For CM, younger adults had higher binocular summation ratio (1.39±0.08) compared to older adults (1.12±0.09). Binocular viewing improved VA with LM letters for both groups similarly. However, in older adults, binocular viewing did not improve VA with CM letters as much as in younger adults. This could reflect a decline of higher visual areas due to ageing process, most likely higher than V1, which may be missed if measured with luminance-based stimuli alone.",
"keywords": [
"Visual acuity",
"first-order",
"second-order",
"extrastriate cortex",
"luminance-modulated",
"contrast-modulated",
"ageing",
"binocular vision"
],
"content": "Introduction\n\nVisual acuity (VA) measurement is one of the clinical routines for ocular examination. VA is the capacity for seeing distinctly the details of an object. It can be described as the eye’s ability to discriminate or resolve spatially organised details. It is usually represented in two ways, which are the reciprocal of the minimum angle of resolution and Snellen fraction. The common acuity charts which are widely used are Snellen chart, logMAR chart, Kay pictures, and etc. Several conditions are known to have an impact on VA, including blur, amblyopia (lazy eye) and normal ageing (e.g. Chung et al., 2007; Elliot et al., 1995). Clinical letter acuity charts rely on the ability of the patient to discriminate between different letters or optotypes. A common feature of letter charts is having black letters on a white background, resulting in maximum difference in luminance or brightness between them. As such, these letters can be classified as luminance-based or first order stimuli.\n\nThe human visual system is sensitive at detecting objects or images irrespective of the types of features defining them. First-order or luminance-defined information is known to be processed by linear mechanisms through linear processing within the striate visual cortex (V1). Stimuli which portray variation in properties such as contrast, texture or orientation without any change in mean luminance are known as second-order stimuli (Hairol et al., 2013; Sukumar & Waugh, 2007; Wong et al., 2001). Processing mechanisms of second-order stimuli are thought to be more complex, and occur in higher and more binocular areas of the visual cortex, than those of first-order stimuli (e.g. Calvert et al., 2005; Hairol & Waugh, 2010; Wong et al., 2005). Neurophysiology studies in cat (Mareschal & Baker, 1998) and primates (Baker & Mareschal, 2001) support the idea that neurons in extrastriate visual cortex, for example V2, are more responsive to contrast-modulated stimuli, compared to neurons in V1. Further evidence is shown by functional magnetic resonance imaging (fMRI) findings in human brain (Ashida et al., 2007; Larsson et al., 2006).\n\nIt is well known that human visual performance reduces with normal ageing, i.e., ageing that is free of pathology or disease. In the past 30 years, many experimental studies have been conducted to investigate changes in a number of visual functions. Frisen & Frisen (1981) stated that VA shows a monotonic rise towards the age of 25 years and a gradual decline thereafter. The most marked decline occurs after the age of 60. Another study showed a slow but significant gradual worsening of VA after 50 years of age (Elliot et al., 1995). Spatial contrast sensitivity for first-order, luminance based gratings is also one of the visual functions which shows significant reduction in elderly compared to younger adults, even when senile miosis and reduced optical transmission factors are taken into consideration (Elliott et al., 1990). Contrast sensitivity for second-order stimuli in healthy elderly declines earlier with slower progression rate compared to that measured with first-order stimuli (Tang & Zhou, 2009). These visual deficits in the elderly cannot be fully attributed to optical changes, but may be due to changes in retina and/or visual pathway (Spear, 1993). Stereoacuity in healthy older adults is also reduced even without cognitive impairment such as Alzheimer’s disease (Bassi et al., 1993), implying that deterioration in binocular vision and binocular neurons of the visual cortex occurs later in life, even when VA is relatively spared. Ageing also increases contrast threshold for detecting second-order stimuli than for first-order stimuli (Habak & Faubert, 2000). As the ageing population increases, there is a pressing need to identify the nature of perceptual capabilities in elderly. In this study, we measured and compared VA between visually healthy older and younger adults using luminance-modulated (LM) and contrast-modulated (CM) noise letters using the staircase method. This study of age-related visual system can act as a model for addressing questions relevant to a general understanding of the effects of ageing on neural information processing and VA deterioration.\n\n\nMethods\n\nFive older adults (mean age: 54.0±1.83 years old) and five younger adults (mean age: 25.4±1.29 years old) were recruited for the experiment. All of them underwent complete ocular health examinations to ensure that no ocular pathologies or binocular anomalies were present. None of them had any history of systemic diseases or medication with known ocular involvement. All participants wore their best refractive correction, with corrected distance VA of logMAR 0.1 (Snellen 6/7.5) or better for older adults and logMAR 0.0 (Snellen 6/6) or better for younger adults. Two sessions of training (approximately 1 hour) was made compulsory before formal data collection began to ensure that participants were familiar with the experiment. Written consent was obtained from all participants before the start of any data collection. The Ethics Committee of Faculty of Health Sciences, Universiti Kebangsaan Malaysia approved the conduct of this research (UKM 1.5.3.5/244/NN-053-2015).\n\nStimuli were displayed on a computer screen (ViewSonic Professional Series P227f) using a custom-written program in Matlab (Mathworks, Inc) on a Dell Precision T1600 CPU. The stimuli were loaded on to the frame store memory of a VSG graphic card (Cambridge Research Systems) installed in the computer. Monitor calibration and gamma correction procedures were carried out every 3 to 6 months by using OptiCal photometer to avoid adjacent pixel nonlinearity (Bertone et al., 2011; Hairol et al., 2013). In every session, the display monitor was turned on for at least 20 minutes to stabilise its luminance output before data collection commenced.\n\nRecognition of luminance-modulated (LM) and contrast-modulated (CM) letters was determined using H, O, T, and V, derived from the clinically used Sloan letters. The HOTV letters were constructed on a 5×5 template, where each stroke of the letter is one fifth of the letter’s size. The LM letters (an example shown in Figure 1a) were created by adding a luminance modulation function to a binary white noise carrier. The CM letters (an example shown in Figure 1b) were created by multiplying a modulation function with a binary white noise carrier (eg. Chung et al., 2006; Hairol et al., 2013; Hairol & Waugh, 2010). The stimuli can be mathematically expressed as follows:\n\nI(x, y) = I [1+ nN(x, y) + lL(x, y) + mnM(x, y)N(x, y)] (Equation 1)\n\nwhere I (x, y) is the luminance at position (x, y); I is the mean luminance; n is the noise contrast, which was fixed at 0.2 for all experiments; N(x, y) is the binary noise value at position (x, y) of −1 or 1; l is the luminance amplitude, which is zero for CM letters; m is the contrast amplitude, which is zero for LM letters; L(x, y) is the luminance modulation function, a square wave; and M(x, y) is the contrast modulation, also a square wave. For generation of LM and CM stimuli, either l or m was adjusted, respectively, the other being set to zero. Total size of noise matrix was 500 pixels. Noise checks were scaled to the letter size and each letter consisted of 15 noise checks with 0.47 mm pixel size for one noise check. Noise was presented dynamically throughout the experiment to avoid any luminance artefacts which may occur due to pixel clumping (Smith & Ledgeway, 1997; Sukumar & Waugh, 2007).\n\nStimuli, (a) luminance-modulated (LM) letter and (b) contrast-modulated (CM) letter.\n\nLetter resolution threshold was measured using staircase method with a four spatial alternative-forced-choice paradigm. This method allows relatively quick estimation of threshold. The two down, one up staircase provided threshold estimation at 70.7% correct (Shen, 2013). Participants recorded what they saw by pressing the appropriate key on the keyboard. After two successive correct responses, the size of the letter was reduced by approximately 0.125 logMAR. An incorrect response resulted in 0.125 logMAR increase in the letter size, i.e. a reversal of the staircase. There was no time limit for stimulus presentation. Eight reversals of staircase method ended the experimental run, and acuity threshold was estimated using the last six reversals. A run consisted of 30–40 trials. Data from four runs were averaged to obtain the mean acuity threshold. The experiment was run under binocular viewing and monocular viewing. In monocular viewing, the non-dominant eye was occluded with a black patch. The viewing distance between participant and monitor screen was 9 m for LM letters (achieved with a front-surfaced mirror) and 4.5 m for CM letters. Room luminance was kept constant across the testing distance.\n\n\nResults\n\nThe mean VA (logMAR) measured with LM and CM letters of all participants are shown in Table 1. For monocular viewing, in younger adults, VA with LM letters was 1.18 × better than that in older adults while VA with CM letters was 1.24 × better than that in older adults. For binocular viewing, VA with LM letters in younger adults was 1.27 × better than that in older adults while VA with CM letters in younger adults was 1.58 × better than that in older adults.\n\nIn Figure 2, VA of younger adults were always better than older adults’, regardless of stimulus type and viewing condition. For monocular viewing, VA with LM letters were significantly better than with CM letters in older and younger adults [F(1, 8)= 427.63, p< 0.001]. There was no significant main effect of age [F(1, 8)= 2.50, p> 0.05] on VA. There was also no significant interaction between stimulus type and age on VA [F(1, 8)= 0.50, p> 0.05], that is, the difference in VA in the two participant groups was similar for the two stimulus types. For binocular viewing, VA with LM letters were significantly better than with CM letters in older and younger adults [F(1, 8)= 609.58, p< 0.001]. However, there was a significant main effect of age on VA [F(1, 8)= 5.67, p< 0.05], that is, the difference in VA in the two participant groups was significantly different for the two stimulus types [F(1, 8)= 7.27, p< 0.05].\n\nTable 2 shows binocular summation ratios [defined as monocular VA (MAR)÷binocular VA (MAR)] for LM and CM letters in older and younger adult groups. ANOVA test showed no main significant effect of age on stimulus types and binocular summation ratios, [F(1, 8)= 0.67, p> 0.05]. However, the mean difference between binocular acuities of CM letters in older and younger adults was 0.19 logMAR, which was approximately two lines worse on the letter charts and may be of clinical significance.\n\nRetinal illuminance for older adults is reduced to between 10%–33% that of younger adults (Weale, 1961). In order to simulate the reduced retinal illumination in older adults, younger participants were tested with neutral density (ND) filters. Three of the younger participants were re-tested with 85N6 ND filters (Kodak) which reduced light transmission to 19%. The stimuli and procedures were the same as described in the main experiment. VA with and without ND filters for LM and CM letters are compared in Figure 3. There was no interaction between stimulus type and retinal illumination on VA for both monocular [F(1, 2)= 1.44, p> 0.05] and binocular viewing [F(1, 2)= 14.85, p> 0.05], that is, the difference in VA for LM and CM letters measured with and without ND filters was not statistically significant.\n\n\nConclusions and discussion\n\nVA with LM and CM letters in visually normal older and younger adults were investigated in this study. In previous psychophysical studies, VA with LM stimuli is found to be better than that with CM stimuli in normal young adults (Hairol et al., 2013; Waugh et al., 2010). Similar results were shown in our study, where VA with LM letters was better than that with CM letters in visually normal older and younger adults (Table 1 and Figure 2). The worse acuity for CM letters suggests that larger scale mechanisms are needed for CM information processing compared to that for LM, similar to the findings of Schofield & Georgeson (1999), and Sukumar & Waugh (2007).\n\nAn effect of ageing on perception of CM stimuli has been reported by Habak & Faubert (2000). They found that the contrast sensitivity for CM stimuli of older adults was significantly worse than that for LM stimuli, which is consistent with our results. Besides, our findings is in accordance with study of Tang & Zhou (2009) as well. They showed that contrast sensitivity for second-order stimuli begins to decline significantly earlier than for first order stimuli, and with a slower rate of progression. These findings suggest that CM stimulus processing mechanisms may be more vulnerable to neurophysiological changes during ageing.\n\nA noteworthy finding in this study is that the binocular summation ratio for CM letters in younger adults was higher than that for LM letters, while the binocular summation ratio for CM letters was almost similar to LM letters in older adults. Waugh & her colleagues (2009) measured the monocular and binocular detection thresholds for LM and CM Gaussian blobs, and showed that binocular summation ratios for CM stimuli were equal or higher than that for LM Gabors for all modulation frequencies above 0.5 cycles per degree (cpd), and were more consistent across modulation frequencies. The findings led to a speculation that CM stimuli are likely to be processed in more binocular areas than LM stimuli, such as in V2, given the predominantly binocular nature of V2 neurons (Hubel & Livingstone, 1987). Indeed, human fMRI study (Calvert et al., 2005) and psychophysical study in amblyopes (Wong et al., 2005) also suggested that CM processing may involve higher visual areas than that for LM. The lesser improvement of CM VA during binocular viewing in older adults compared to younger adults, suggests that CM stimuli neural processing mechanisms may be more vulnerable to neurophysiological changes that are associated with increasing age, may start earlier in higher visual areas. In fact, single-unit recordings showed decreases in which signal-to-noise ratio and sensitivity in cortical neurons of elderly monkeys, and these losses were even more robust in V2 than in V1 neurons (Wang et al., 2005).\n\nOur results of binocular summation ratios for LM stimuli in younger and older adults are not consistent with the findings of Pardhan (1996). She measured monocular and binocular contrast sensitivities at spatial frequencies of one and six cpd in young and older adults with normal healthy eyes. Binocular summation ratios were higher for the younger adults compared to the older adults for both spatial frequencies. However, vertical sinusoidal gratings were used as the target for her study for determining participants’ contrast sensitivity threshold while we used noise letter stimuli for resolution threshold. The difference in binocular summation ratios between younger and older adults in Pardhan’s study could be due to the difference in stimuli and tasks.\n\nWe have shown that VA reduction in magnitude for CM letters was greater than that for LM letters under binocular viewing. However, older adults experience age-related physiological changes in vision, which included senile miosis, ie. reduction in pupil size. Senile miosis lowers retinal illumination levels, thereby affecting visual performance in elderly (Winn et al., 1994). Reduced retinal illumination in older adults may have lead to the difference between older and younger participants’ visual performance. Our retinal illumination control experiment showed no significant difference between VA for both stimulus types with and without the presence of neutral density filters. This is consistent with the study done by Habak & Faubert (2000) where they showed that contrast thresholds for LM and CM gratings are similar with or without neutral density filters on younger adults. Therefore, it appears unlikely that our findings were a result of the reduced retinal illumination in our older adults group.\n\nIn conclusion, reduction of VA for CM letters is higher than that for LM letters especially in binocular vision for healthy older adults compared to younger adults. This suggests that a young and intact visual cortex plays a key role for good visual performance with CM letters. This extra age-related VA deficit may not be fully revealed when measured with luminance-based stimuli alone. A decrease in binocular summation ratio with ageing for CM letters in older adults may reflect an early decline in higher visual areas, most likely higher than V1. This speculation is supported by the study of Costa et al. (2013) which suggested that ageing might have a more pronounced effect in higher visual areas than in the primary striate cortex. It is vital to know what biological (e.g. neural function, optical changes) and environmental characteristics (e.g. daily lifestyle, dietary) differentiate those older adults who lose little to no visual performance as they age and those who do not (Owsley, 2011). It is also prudent to include a clinical measure of VA to examine age-associated differences in neural activity during cognitive processing that may affect neurologic functioning, e.g. dementia, cerebrovascular disease and depression (Daffner et al., 2013). A quick yet effective VA test is commonly being relied on to measure visual performance. However, the rate of VA changes during ageing which measure with luminance-based acuity charts may not be sufficient to be considered as ‘clinically normal’. Therefore, the advantage of CM stimuli which may serve to more sensitively detect early visual deterioration in ageing should be further investigated. A limitation in our study is that our older adults (54.0±1.83 years old) are relatively young by WHO definition (http://www.who.int/healthinfo/survey/ageingdefnolder/en/). Therefore, a model of VA deterioration with LM and CM stimuli throughout wider normal healthy age groups is worth to be explored in future studies, which could lead to the development of a prototype of a novel CM based acuity test that might benefit the older age population.\n\n\nData availability\n\nF1000Research: Dataset 1. VA with LM and CM letters in older and younger adults, 10.5256/f1000research.9410.d132585 (Woi et al., 2016a).\n\nF1000Research: Dataset 2. Binocular summation ratios with LM and CM letters in older and younger adults, 10.5256/f1000research.9410.d132586 (Woi et al., 2016b).\n\nF1000Research: Dataset 3. VA for both stimulus types with and without neutral density (ND) filter in younger adults, 10.5256/f1000research.9410.d132587 (Woi et al., 2016c).",
"appendix": "Author contributions\n\n\n\nSJW, MIH, SK and WPJ conceived the study. MIH and PJW designed the experiments. PJW carried out the research. MIH, SK and SJW contributed to the design of experiments and provided expertise. PJW prepared the first draft of the manuscript. MIH contributed to the experimental design and preparation of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the Universiti Kebangsaan Malaysia grant (GUP-2015-049).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors are grateful to the participants for their commitment and cooperation.\n\n\nReferences\n\nAshida H, Lingnau A, Wall MB, et al.: fMRI Adaptation Reveals Separate Mechanisms for First-Order and Second-Order Motion. J Neurophysiol. 2007; 97(2): 1319–1325. PubMed Abstract | Publisher Full Text\n\nBaker CL, Mareschal I: Processing of second-order stimuli in the visual cortex. Prog Brain Res. 2001; 134: 171–191. PubMed Abstract | Publisher Full Text\n\nBassi CJ, Solomon K, Young D: Vision in aging and dementia. Optom Vis Sci. 1993; 70(10): 809–813. PubMed Abstract | Publisher Full Text\n\nBertone A, Guy J, Faubert J: Assessing spatial perception in aging using an adapted Landolt-C technique. Neuroreport. 2011; 22(18): 951–955. PubMed Abstract | Publisher Full Text\n\nCalvert J, Manahilov V, Simpson WA, et al.: Human cortical responses to contrast modulations of visual noise. Vision Res. 2005; 45(17): 2218–30. PubMed Abstract | Publisher Full Text\n\nChung ST, Jarvis SH, Cheung S: The effect of dioptric blur on reading performance. Vision Res. 2007; 47(12): 1584–1594. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChung ST, Li RW, Levi DM: Identification of contrast-defined letters benefits from perceptual learning in adults with amblyopia. Vision Res. 2006; 46(22): 3853–3861. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCosta TL, Nogueira RM, Pereira AG, et al.: Differential effects of aging on spatial contrast sensitivity to linear and polar sine-wave gratings. Braz J Med Biol Res. 2013; 46(10): 855–860. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDaffner KR, Haring AE, Alperin BR, et al.: The Impact of Visual Acuity on Age-Related Differences in Neural Markers of Early Visual Processing. Neuroimage. 2013; 67: 127–136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElliot DB, Yang KC, Whitaker D: Visual acuity changes throughout adulthood in normal, healthy eyes: seeing beyond 6/6. Optom Vis Sci. 1995; 72(3): 186–191. PubMed Abstract | Publisher Full Text\n\nElliott D, Whitaker D, MacVeigh D: Neural contribution to spatiotemporal contrast sensitivity decline in healthy ageing eyes. Vision Res. 1990; 30(4): 541–547. PubMed Abstract | Publisher Full Text\n\nFrisén L, Frisén M: How Good is Normal Visual Acuity? A study of letter acuity thresholds as a function of age. Albrecht Von Graefes Arch Klin Exp Ophthalmol. 1981; 215(3): 149–157. PubMed Abstract | Publisher Full Text\n\nHabak C, Faubert J: Larger effect of aging on the perception of higher-order stimuli. Vision Res. 2000; 40(8): 943–950. PubMed Abstract | Publisher Full Text\n\nHairol MI, Formankiewicz MA, Waugh SJ: Foveal visual acuity is worse and shows stronger contour interaction effects for contrast-modulated than luminance-modulated Cs. Vis Neurosci. 2013; 30(3): 105–120. PubMed Abstract | Publisher Full Text\n\nHairol MI, Waugh SJ: Lateral facilitation revealed dichoptically for luminance-modulated and contrast-modulated stimuli. Vision Res. 2010; 50(23): 2530–2542. PubMed Abstract | Publisher Full Text\n\nHubel DH, Livingstone MS: Segregation of form, color, and stereopsis in primate area 18. J Neurosci. 1987; 7(11): 3378–3415. PubMed Abstract\n\nLarsson J, Landy MS, Heeger DJ: Orientation-selective adaptation to first- and second-order patterns in human visual cortex. J Neurophysiol. 2006; 95: 862–881. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMareschal I, Baker CL Jr: Temporal and spatial response to second-order stimuli in cat area 18. J Neurophysiol. 1998; 80(6): 2811–2823. PubMed Abstract\n\nOwsley C: Aging and vision. Vision Res. 2011; 51(13): 1610–1622. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPardhan S: A comparison of binocular summation in young and older patients. Curr Eye Res. 1996; 15(3): 315–9. PubMed Abstract | Publisher Full Text\n\nSchofield AJ, Georgeson MA: Sensitivity to modulations of luminance and contrast in visual white noise: separate mechanisms with similar behaviour. Vision Res. 1999; 39(16): 2697–2716. PubMed Abstract | Publisher Full Text\n\nShen Y: Comparing adaptive procedures for estimating the psychometric function for an auditory gap detection task. Atten Percept Psychophys. 2013; 75(4): 771–780. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith AT, Ledgeway T: Separate detection of moving luminance and contrast modulations: fact or artifact? Vision Res. 1997; 37(1): 45–62. PubMed Abstract | Publisher Full Text\n\nSpear PD: Neural bases of visual deficits during aging. Vision Res. 1993; 33(18): 2589–2609. PubMed Abstract | Publisher Full Text\n\nSukumar S, Waugh SJ: Separate first- and second-order processing is supported by spatial summation estimates at the fovea and eccentrically. Vision Res. 2007; 47(5): 581–596. PubMed Abstract | Publisher Full Text\n\nTang Y, Zhou Y: Age-related decline of contrast sensitivity for second-order stimuli: earlier onset, but slower progression, than for first-order stimuli. J Vis. 2009; 9(7): 18. PubMed Abstract | Publisher Full Text\n\nWang Y, Zhou Y, Ma Y, et al.: Degradation of signal timing in cortical areas V1 and V2 of senescent monkeys. Cereb Cortex. 2005; 15(4): 403–8. PubMed Abstract | Publisher Full Text\n\nWaugh SJ, Formankiewicz MA, Ahmad N, et al.: Effects of dioptric blur on foveal acuity and contour interaction for noisy Cs. J Vis. 2010; 10(7): 1330. Publisher Full Text\n\nWaugh SJ, Lalor SJ, Hairol MI: Binocular summation for luminance- and contrast-modulated noise stimuli. J Vis. 2009; 9(8): 1012. Publisher Full Text\n\nWeale RA: Retinal Illumination and Age. Trans. Illum. Eng. Soc., Light Res Technol. 1961; 26(2): 95–100. Publisher Full Text\n\nWinn B, Whitaker D, Elliott DB, et al.: Factors Affecting Light-Adapted Pupil Size in Normal Human Subjects. Invest Ophthalmol Vis Sci. 1994; 35(3): 1132–1137. PubMed Abstract\n\nWoi PJ, Hairol MI, Waugh SJ, et al.: Dataset 1 in: Visual acuity measured with luminance-modulated and contrast-modulated noise letter stimuli in young adults and adults above 50 years old. F1000Research. 2016a. Data Source\n\nWoi PJ, Hairol MI, Waugh SJ, et al.: Dataset 2 in: Visual acuity measured with luminance-modulated and contrast-modulated noise letter stimuli in young adults and adults above 50 years old. F1000Research. 2016b. Data Source\n\nWoi PJ, Hairol MI, Waugh SJ, et al.: Dataset 3 in: Visual acuity measured with luminance-modulated and contrast-modulated noise letter stimuli in young adults and adults above 50 years old. F1000Research. 2016c. Data Source\n\nWong EH, Levi DM, McGraw PV: Spatial interactions reveal inhibitory cortical networks in human amblyopia. Vision Res. 2005; 45(21): 2810–9. PubMed Abstract | Publisher Full Text\n\nWong EH, Levi DM, McGraw PV: Is second-order spatial loss in amblyopia explained by the loss of first-order spatial input? Vision Res. 2001; 41(23): 2951–60. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15649",
"date": "25 Aug 2016",
"name": "Goro Maehara",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors examined the effect of aging on visual acuity measured using the luminance-modulated and contrast-modulated letter stimuli. The main claim of the paper is that the advantage of binocular viewing (binocular summation) was smaller for the older adults than for the younger adults when the contrast-modulated stimuli were presented. Although their aim is of scientific and clinical interests, the results are not clear enough to draw any strong conclusion. Comments are detailed below.\n\nMajor Concerns:\n1. The main experiment\nVisual acuity was generally better for binocular viewing than for monocular viewing. The younger adults showed better visual acuity than the older adults. These results are clear but not surprising. Although the interactions among age, stimulus types, and viewing conditions are important here, they were too small to draw any conclusions. The authors reported that there were the significant interactions. However, I guess that they conducted the two-way ANOVAs separately for binocular and monocular viewing. This is not appropriate. The authors should conduct the three-way (age × stimulus type × viewing condition) ANOVA. This ANOVA yielded no significant interaction. I encourage the authors to increase the number of observers for making meaningful conclusions because there were only 5 observers for each age group.\n2. The additional experiment\nIt can be seen from Figure 1 and 3 that visual acuity was comparable between the older adults and the younger adults who viewed the stimuli through a ND filter. This suggests that the reduction in retinal illumination caused the differences between the age groups. Please analyze and discuss this comparison. The authors argued that the age-group differences cannot be attributed to the illumination reduction because there was no significant interaction between stimulus type and retinal illumination. However, only 3 observers participated in the experiment. Their argument is far from convincing. It is necessary to increase the number of observers. Please conduct the three-way ANOVA for this experiment, too.\n\nMinor concerns:\n1. Please state clearly what statistical analyses were conducted.\n\n2. Typo in page 3. “artcfacts” should be “artifacts”.",
"responses": []
},
{
"id": "16356",
"date": "16 Sep 2016",
"name": "Wan-Hazabbah Wan Hitam",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall, this is a good study adding a new knowledge to the field.\n\nComment and suggestions:\nThe title is very interesting.\n\nThe abstract does not contain clear objective of the article written. The conclusion also is not clearly stated.\n\nThere are no keywords written.\n\nThe introduction is good and the objective is stated nicely.\n\nThe methods section is overall acceptable. However, the sample size calculation is not mentioned. The number of sample collected is rather small.\n\nThe results presented are satisfactory. Analysis done is appropriate.\n\nOverall the discussion is good. The limitation of the study should be addressed clearly.\n\nThe conclusion is too long. It should answer the objective.\n\nThe number of references is adequate and updated\n\nThe Figures and Tables are satisfactory.",
"responses": []
},
{
"id": "18036",
"date": "28 Nov 2016",
"name": "Li-Ting Tsai",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well written manuscript describing the effects of aging on monocular and binocular visual acuity by recognizing luminance-modulated and contrast-modulated letters. There are some points, outlined below, which need clarification or further consideration by the authors.\nThe visual acuity (Methods) for younger adults seems to be better than the older adults. Could the authors add each observer’s original data for OD, OS, and OU acuity as another supplementary data? And note which eye was used in the monocular experiment?\n\nIn the method session, please explain the experimental sequences of LM and CM under monocular and binocular viewing.\n\nIn the method session, please add the value of contrast to display the letter stimuli.\n\nThe additional experiment - the authors tended to use the ND filters to simulate the reduced retinal illumination in older adults due to senile miosis. However, the possible reasons or mechanism related vision changes in older adult are not limited to senile miosis or higher visual areas. In fact, the visual acuity of older adults seems to be lower than those of younger group. However, it is suggested to discuss other possible reasons to contribute the results.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1961
|
https://f1000research.com/articles/5-1949/v1
|
11 Aug 16
|
{
"type": "Method Article",
"title": "Rapid and high throughput molecular identification of diverse mosquito species by high resolution melting analysis",
"authors": [
"Yvonne Ukamaka Ajamma",
"Enock Mararo",
"David Omondi",
"Thomas Onchuru",
"Anne W. T. Muigai",
"Daniel K Masiga",
"Jandouwe Villinger",
"Yvonne Ukamaka Ajamma",
"Enock Mararo",
"David Omondi",
"Thomas Onchuru",
"Anne W. T. Muigai",
"Daniel K Masiga"
],
"abstract": "Mosquitoes are a diverse group of invertebrates, with members that are among the most important vectors of diseases. The correct identification of mosquitoes is paramount to the control of the diseases that they transmit. However, morphological techniques depend on the quality of the specimen and often unavailable taxonomic expertise, which may still not be able to distinguish mosquitoes among species complexes (sibling and cryptic species). High resolution melting (HRM) analyses, a closed-tube, post-polymerase chain reaction (PCR) method used to identify variations in nucleic acid sequences, has been used to differentiate species within the Anopheles gambiae and Culex pipiens complexes. We validated the use of PCR-HRM analyses to differentiate species within Anopheles and within each of six genera of culicine mosquitoes, comparing primers targeting cytochrome b (cyt b), NADH dehydrogenase subunit 1 (ND1), intergenic spacer region (IGS) and cytochrome c oxidase subunit 1 (COI) gene regions. HRM analyses of amplicons from all the six primer pairs successfully differentiated two or more mosquito species within one or more genera (Aedes (Ae. vittatus from Ae. metallicus), Culex (Cx. tenagius from Cx. antennatus, Cx. neavei from Cx. duttoni, cryptic Cx. pipiens species), Anopheles (An. gambiae s.s. from An. arabiensis) and Mansonia (Ma. africana from Ma. uniformis)) based on their HRM profiles. However, PCR-HRM could not distinguish between species within Aedeomyia (Ad. africana and Ad. furfurea), Mimomyia (Mi. hispida and Mi. splendens) and Coquillettidia (Cq. aurites, Cq. chrysosoma, Cq. fuscopennata, Cq. metallica, Cq. microannulatus, Cq. pseudoconopas and Cq. versicolor) genera using any of the primers. The IGS and COI barcode region primers gave the best and most definitive separation of mosquito species among anopheline and culicine mosquito genera, respectively, while the other markers may serve to confirm identifications of closely related sub-species. This approach can be employed for rapid identification of mosquitoes.",
"keywords": [
"High resolution melting analysis",
"molecular identification",
"mosquitoes",
"Aedes",
"Culex",
"Mansonia",
"Anopheles"
],
"content": "Introduction\n\nMosquitoes are among the most important disease vectors, known to transmit and maintain the circulation of pathogens that cause both global and neglected tropical diseases in humans and animals1. The correct identification of different field-collected mosquito species, endemic to distinct ecologies, with high parasite and arthropod-borne virus (arbovirus) diversities is crucial to the planning of targeted vector control strategies to mitigate disease transmission2. The last and most comprehensive Afrotropical mosquito identification keys were published in 1941 for culicines3 and in 1987 for anophelines4. Molecular approaches that efficiently differentiate conspecific mosquitoes such as the barcode region5 improve identification accuracy considerably6, but are time consuming, expensive in terms of post-polymerase chain reaction (post-PCR) processing and depend heavily on DNA sequencing.\n\nRecent approaches have taken advantage of the unique melting profiles generated by homologous PCR products with small sequence differences during high resolution melting (HRM) analysis7,8. Indeed, PCR-HRM has been used to differentiate mosquito transmitted arboviruses9–11 and malaria Plasmodium12,13, vertebrate blood meals of mosquitoes10, between two members of the Anopheles gambiae complex14 and amongst three members of the Culex pipiens complex15. HRM analysis has proven to offer higher resolution of PCR product based species identification on sequence variants than electrophoretic methods by revealing even single nucleotide polymorphisms (SNPs) in the simple sequence repeats (SSRs) among products of similar sizes16,17. Conventional detection of specific PCR products sequence relies on costly molecular probes and/or product sequencing18. For species identification16, only representative samples with distinct HRM profiles need to be sequenced, thereby reducing reagent and sample consumption costs10–11. Combining HRM analysis of barcode region sequences (Bar-HRM) has been successfully used to rapidly and accurately distinguish between closely related antelope species19 and medicinal plants20,21 and to authenticate the source of vegetable oils22.\n\nAlthough HRM has been successfully used to differentiate between specific Anopheles and Culex mosquitoes, the approach’s broader applicability and most suitable markers have not been evaluated. Previously, only the ribosomal DNA was targeted for An. gambiae sensu lato (s.l.)14 and only the acetylcholinesterase gene was used in distinguishing the Cx. pipiens complex15. This study aimed at validating the use of HRM analysis for high throughput molecular culicine and anopheline mosquito identification and differentiation, comparing the utility of one ribosomal IGS (previously used to differentiate An. gambiae s.l.)14 and three mitochondrial (COI, ND1, cyt b) gene markers.\n\n\nMethods\n\nWe used 109 mosquitoes (Table 1 and Table 2) that were collected in 2012 during the rainy seasons near Lake Baringo from March 2–4, July 16–24 and October 12–21 and Lake Victoria from April 2–15, May 18–31 and November 12–29 during a mosquito diversity study around the islands and mainland shores of Lake Baringo in Baringo County (Table 1) and Lake Victoria in Homa Bay County (Table 2) in Kenya6. Before sampling, we obtained ethical clearance for the study from the Kenya Medical Research Institute (KEMRI) ethics review committee (Approval Ref: Non-SSC Protocol #310). These mosquitoes were morphologically identified during a previous study6. Baringo County is a known hotspot for arbovirus outbreaks23, while Homa Bay County is endemic to malaria and is located in a region with a history of arbovirus activity10. One sample each of Anopheles gambiae sensu stricto (s.s.) and An. arabiensis, Aedes aegypti and Culex pipiens from laboratory colonies maintained in the Insectary of the International Centre of Insect Physiology and Ecology (icipe), Nairobi, Kenya, were used as controls. Also, specimens with confirmed identity that have been previously sequenced and submitted to GenBank (Table 1 and Table 2) were used as both controls and samples.\n\nGenBank accessions are provided only for samples with confirmed identity and from which the COI DNA sequences were obtained during a previously published mosquito diversity study6.\n\nGenBank accessions are provided only for samples with confirmed identity and from which the COI DNA sequences were obtained during a previously published mosquito diversity study6.\n\nFrom each mosquito, we extracted DNA according to the hot sodium hydroxide and Tris (HotSHOT) DNA extraction protocol24 from a single mosquito leg that was detached from the rest of the body using a pair of forceps and dissecting pin. Without crushing, the mosquito leg was put in a 0.2 ml microcentrifuge tube containing 30 µl of Alkaline Lysis buffer (25 mM NaOH (Thermo Fisher Scientific, Pittsburgh, USA), 0.2 mM disodium EDTA (Thermo Fisher Scientific), pH 8.0) and incubated in a thermocycler at 95°C for 30 minutes and cooled at 4°C for 5 minutes. Then, 30 µl neutralising solution (40 mM Tris-HCl (Thermo Fisher Scientific)) was added. The resulting DNA was stored at -20°C until required as templates for PCR assays.\n\nBased on multiple alignments using Geneious software version 8.1.425 of mitochondrial genomes of mosquitoes (GenBank accessions NC_015079, NC_028616, NC_028223, KR068634, NC_010241, NC_014574, EU352212, NC_008070, KT358413, KT382816, KU494979, JX040513, AY729979, KU494979), we designed four sets of primers from two mitochondrial gene regions: COI (COI-AnophF/HCO2108R; Uni-Minibar-JVF/Uni-Minibar-JVR; Mos-CO1-JVF/Mos-CO1-JVR) and ND1 (Mos-ND1F/Mos-ND1R) genes (Table 3). The COI AnophF primer was initially designed specifically for Anopheles mosquitoes to be used with the HCO2108R primer26, but tested on other species as well. Using samples of morphologically and molecularly identified Culex, Aedeomyia, Mimomyia, Coquillettidia, Mansonia, Aedes, and Anopheles mosquito species (Table 1 and Table 2), we amplified different gene regions of their genomes using six pairs of primers (Table 3) in three replicate runs of single-plex PCRs in a Rotor-Gene Q HRM real time PCR thermocycler (QIAGEN, Hannover, Germany). PCR grade water was used as negative control while mosquito species from Ae. aegypti, An. gambiae s.s., An. arabiensis and Cx. Pipiens quinquefasciatus colonies maintained in the International Centre of Insect Physiology and Ecology (icipe) Insectary Unit were used as positive controls. The PCR mix contained 5 µl of 5X Hot Firepol EvaGreen HRM Mix (Solis BioDyne, Tartu, Estonia), 0.5 µM of each primer, 1 µl of DNA template and distilled water in a final volume of 10 µl. The thermal cycling conditions involved an initial denaturation for 1 minute at 95°C, followed by 35 cycles of denaturation at 95°C for 30 seconds, annealing at 50°C for 20 seconds, and extension at 72°C for 30 seconds, and a final extension at 72°C for 7 minutes. Without stopping the reaction, the PCR amplicons were denatured at 95°C for 1 minute, held for another minute at 40°C and melted by gradually raising the temperature from 70°C to 95°C by 0.1°C in 2 second steps, waiting for 90 seconds of pre-melt conditioning on first step and 2 seconds in subsequent steps. The outcome was automatically plotted on a connected computer and visually observed and analysed using the Rotor-Gene Q Series software v2.1. Representative samples of differentiated mosquito species that had similar HRM curves were purified with ExoSAP-IT (USB Corporation, Cleveland, OH) and submitted for DNA sequencing at Macrogen (South Korea). To confirm the identity of PCR-HRM differentiated mosquitoes, DNA sequences were edited with Geneious version 8.1.425 and queried against the GenBank nr database (http://www.ncbi.nlm.nih.gov/) using the Basic Local Alignment Search Tool (BLAST N) version 2.3.027.\n\nF is forward primer direction; R is reverse primer direction.\n\n\nResults\n\nWe differentiated 12 mosquito species in the Aedes (two), Anopheles (two), Culex (six), and Mansonia (two) genera by HRM analyses (Table 4). The COI sequences of some of the mosquito samples analyzed and differentiated were obtained during a previously published mosquito diversity study6 and their respective GenBank Accession numbers are listed in Table 1 and Table 2. Despite the fact that the COI-AnophF/HCO2198R primers were originally designed based on Anopheles mitochondria genome alignments, they were most efficient in differentiating among Mansonia (Ma. africana and Ma. uniformis (Figure 1A)), Culex (Cx. neavei and Cx. duttoni, Cx. tenagius and Cx. antennatus, and two genetic variants of Cx. pipiens (Figure 2A)), and Aedes (Ae. vittatus and Ae. metallicus (Figure 3)) mosquitoes (Table 4). Indeed, the DNA sequences flanked by the COI-AnophF/HCO2198R primers included multiple polymorphic sites in species within these genera (Figure 4). Although there are SNPs within species DNA that resulted to the slight changes observed in their HRM profiles, the SNPs across species were enough to distinguish between them.\n\nDNS means did not separate. COI means cytochrome c oxidase subunit 1. cyt b means cytochrome B. ND1 means NADH dehydrogenase subunit 1. IGS means intergenic spacer region.\n\nMansonia uniformis and Ma. africana mosquitoes were differentiated by PCR-HRM using the (A) COI-AnophF/HCO2198R, (B) MOS-CO1 and (C) CYT B primer pairs.\n\nCulex species were differentiated by PCR-HRM using the (A) COI-AnophF/HCO2198R, (B) CYT B, (C) Uni-Minibar-JV, and (D) Mos-ND1 primer pairs.\n\nAedes vittatus and Ae. metallicus were differentiated by PCR-HRM using the COI-AnophF/HCO2198R primer pair.\n\nPolymorphic sites vary more between than within species.\n\nMansonia africana and Ma. uniformis could also be differentiated by Mos-COI-JV (Figure 1B) and CYT B (Figure 1C) PCR-HRM analysis. Some Culex species were similarly differentiated by HRM based on their CYT B, Uni-Minibar-JV and Mos-ND1 (Figure 2B–D) primer pair PCR products. The morphologically indistinguishable Cx. tenagius and Cx. antennatus were distinguished only by the COI-AnophF/HCO2198R, CYT B and ND1 primers (Figure 2A, B and D). Similarly, HRM analysis of only two of the COI (COI-AnophF/HCO2198R and Uni-Minibar JV) and the ND1 primer pairs grouped morphologically identical and difficult to differentiate Cx. pipiens into two distinct clades: one with Cx. pipiens voucher sequences from GenBank (KF919189) and those with a sequence that we identified as Culex sp. GPA6 (GenBank accessions KU380352, KU380455, KU380394) (Figure 2A, C and D; Table 4). However, unlike the COI HRM profiles (Figure 2A, B), the ND1 HRM profiles (Figure 2D) of Cx. pipiens amplicons showed a melting temperature shift of to the right (higher temperature) compared to the Culex sp. GPA amplicons, possibly due to greater GC richness of Cx. pipiens at this locus28. Similarly, the IGS primers (AgamUni) differentiated Anopheles gambiae s.s. from An. arabiensis (Figure 5). In addition, the COI-AnophF/HCO2198R primers were also used to separate Cx. neavei from Cx. duttoni (Figure 2A), which belong to the same subgenus of Culex mosquitoes.\n\nTwo sibling species of Anopheles gambiae s.l. were differentiated by PCR-HRM using the AgamUni primer pair.\n\nHRM analysis of all the six primer pairs could not differentiate Aedeomyia (Ad. africana and Ad. furfurea), Mimomyia (Mi. hispida and Mi. splendens) and Coquillettidia (Cq. aurites, Cq. chrysosoma, Cq. fuscopennata, Cq. metallica, Cq. microannulatus, Cq. pseudoconopas and Cq. versicolor) species (Table 4) or among An. funestus and An. coustani species complexes.\n\n\nDiscussion\n\nWe compared six pairs of primers for their potential to differentiate at least two morphologically similar mosquito species within each of seven mosquito genera by PCR-HRM analysis and identified suitable markers for differentiating species within Anopheles, Aedes, Culex and Mansonia mosquitoes. However, none of the markers were suitable for HRM analysis to distinguish among species of Aedeomyia, Mimomyia or Coquillettidia genera mosquitoes. Also, Cx. watti, which can be misidentified morphologically as Cx. duttoni or Cx. pipiens, could not be differentiated by PCR-HRM analyses. Nonetheless, we were able to distinguish Ma. africana from Ma. uniformis, An. gambiae s.s. from An. arabiensis (sibling species of An. gambiae s.l.), Ae. vittatus from Ae. metallicus, as well as Cx. neavei from Cx. duttoni, Cx. tenagius from Cx. antennatus and two cryptic sympatric species of morphologically identical Cx. pipiens. Most notably, the two Cx. pipiens species with distinct COI barcode sequences6 were indeed first identified by HRM analysis of numerous samples6. Thus, the relative economy of HRM analysis compared to sequencing facilitates the rapid identification of cryptic species.\n\nSurprisingly, HRM analysis of PCR products from the COI-AnophF/HCO2198R primers, which were designed for Anopheles, could not distinguish between these sibling species, yet were most effective in discriminating species within the Mansonia, Aedes and Culex genera, including between the cryptic Culex pipiens species. Anopheles gambiae and An. arabiensis were only distinguished using the IGS gene, which was also designed for An. gambiae2 and is routinely used for distinguishing these sibling species by conventional PCR29 and HRM analysis14. In contrast, species complexes of An. coustani and An. funestus were not differentiated with any of the primers. The data suggest that COI30, cyt b and ND1 loci may be unsuitable for distinguishing among Anopheles sibling species. Similarly, the Aedes species could only be differentiated by the COI-AnophF/HCO2198R primers. This could be as a result of more recent speciation, insufficient to allow for sibling species resolution at these markers. Such scenarios have been observed for recent or rapidly evolving groups, such as the Cichlid fishes of eastern Africa, where mitochondrial divergence is not concordant with morphological variations31.\n\nIn contrast, Ma. africana and Ma. uniformis were separated by the COI and cyt b loci, but not by the ND1 and IGS gene primers and Culex species were variably differentiable by all markers, except IGS. For both Mansonia and Culex, as with Aedes, the COI-AnophF/HCO2198R primers were most sensitive in discriminating morphologically indistinct species. This highlights the power of the COI barcode region for identifying diverse cryptic species32. The SNPs present in the COI genes of the ten mosquito species confirms that the COI gene is conserved and polymorphic enough to differentiate these species even in cases of morphological misidentification. The polymorphisms between species were enough to robustly separate them based on their HRM profiles, while sequence polymorphisms within species were too few to significantly alter their HRM profiles.\n\nWe, therefore, recommend the initial use of the COI-AnophF/HCO2198R primers Bar-HRM to differentiate Mansonia, Culex and Aedes mosquito species and the IGS primers for anopheline mosquito identification2,14,33 by HRM. The inability of all the six primer pairs to differentiate many mosquito species among all seven genera tested is an indication that the genetic diversity of many mosquito species is complicated and still poorly understood. Also, the number (sample size) of many of the analyzed mosquito species was small (<3) because these species were scarcely present in the study areas. More samples (≥3) should be used and more study areas should be sampled in subsequent studies to test genetic differentiation of mosquito species34. Additional polymorphic DNA loci should also be identified, tested and used in combination with existing ones for the identification of mosquito species, especially among species complexes and across genera.\n\n\nConclusions\n\nThis study shows that specific PCR markers can be used to distinguish closely related species of mosquitoes using HRM analysis. We distinguished two sibling species of An. gambiae s.l., two species each of Mansonia and Aedes, and six species, including cryptic species, of Culex using six pairs of primers targeting the mitochondrial and ribosomal genes. HRM is a low cost (<$1 per reaction), effective tool that enhances culicine and anopheline mosquito identification and may also reveal population differences in conserved mitochondrial sequences. This approach can improve vector surveillance associated with Plasmodium (malaria) or arbovirus transmission and inform targeted vector control strategies.\n\n\nData availability\n\nAll sequence data associated with this manuscript are freely available in GenBank. All relevant accession numbers are listed in Table 1 and Table 2.\n\nF1000Research: Dataset 1. Raw Rotor-Gene Q HRM data files (.rex), viewable using Rotor-Gene Q software (Qiagen), 10.5256/f1000research.9224.d13056536",
"appendix": "Author contributions\n\n\n\nYUA, DM, JV, and AM conceived of, designed and coordinated the study. YUA, DO and TOO did sample collection and identification. YUA and EM carried out the molecular genetic studies. YUA and JV carried out the sequence analyses and drafted the manuscript. DM, JV and YUA contributed materials used for the study. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe gratefully acknowledge the financial support for this research by the following organizations and agencies: Swedish International Development Cooperation Agency (SIDA), grant number 75000529 to YUA as an African Regional Postgraduate Programme in Insect Science (ARPPIS) student; Funds from Training Health Researchers into Vocational Excellence (THRiVE) in East Africa (grant number 087540) funded by Wellcome Trust to JV and DM supported part of the field sampling. We also acknowledge funding from UK’s Department for International Development (DFID); the Swiss Agency for Development and Cooperation (SDC); and the Kenyan Government.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Laban Njoroge of National Museums of Kenya for helping with morphological identifications of mosquitoes. We acknowledge John Tilion of Ruko Conservancy in Baringo County and Phillip Ojunju of Rusinga Island in Homa Bay County, for helping with mosquito sample collection in the two study areas respectively. We acknowledge the support of Milcah Gitau of icipe’s Insectary Unit in providing the Ae. aegypti, Cx. pipiens, An. gambiae s.s. and An. arabiensis controls. We also thank Esther Waweru of icipe’s Molecular Biology and Bioinformatics Unit (MBBU), Gerard Ronoh, Caroline Tigoi and Geoffrey Jagero of icipe’s ML-EID Laboratory, Lillian Igweta, Lisa Omondi and Margaret Ochanda icipe’s of Capacity Building & Institutional Development (CB&ID) Unit for assisting with logistics.\n\n\nReferences\n\nWeaver SC, Reisen WK: Present and future arboviral threats. Antiviral Res. 2010; 85(2): 328–345. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalker ED, Thibault AR, Thelen AP, et al.: Identification of field caught Anopheles gambiae s.s. and Anopheles arabiensis by TaqMan single nucleotide polymorphism genotyping. Malar J. 2007; 6: 23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards FW: Mosquitoes of the Ethiopian Region. III.- Culicine adults and pupae. British Museum (Natural History), London; 1941. Reference Source\n\nGillies MT, Coetzee M: A Supplement to the Anophelinae of Africa South of the Sahara. Publications of the South African Institute for Medical Research. 1987; (55): 1–143. Reference Source\n\nHebert PD, Cywinska A, Ball SL, et al.: Biological identifications through DNA barcodes. Proc Biol Sci. 2003; 270(1512): 313–321. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAjamma YU, Villinger J, Salifu D, et al.: Abundance and species composition of mosquito vectors of arboviruses in the Lake Victoria and Lake Baringo regions of Kenya. J Med Entomol. 2016. In Press.\n\nRirie KM, Rasmussen RP, Wittwer CT: Product differentiation by analysis of DNA melting curves during the polymerase chain reaction. Anal Biochem. 1997; 245(2): 154–160. PubMed Abstract | Publisher Full Text\n\nVossen RH, Aten E, Roos A, et al.: High-resolution melting analysis (HRMA): more than just sequence variant screening. Hum Mutat. 2009; 30(6): 860–866. PubMed Abstract | Publisher Full Text\n\nNaze F, Le Roux K, Schuffenecker I, et al.: Simultaneous detection and quantitation of Chikungunya, dengue and West Nile viruses by multiplex RT-PCR assays and dengue virus typing using high resolution melting. J Virol Methods. 2009; 162(1–2): 1–7. PubMed Abstract | Publisher Full Text\n\nOmondi D, Masiga DK, Ajamma YU, et al.: Unraveling Host-Vector-Arbovirus Interactions by Two-Gene High Resolution Melting Mosquito Bloodmeal Analysis in a Kenyan Wildlife-Livestock Interface. PLoS One. 2015; 10(7): e0134375. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVillinger J, Mbaya MK, Ouso DO, et al.: Arbovirus and insect-specific virus discovery in Kenya by novel six genera multiplex high resolution melting analysis. Mol Ecol Resour. 2016. In Press.\n\nKipanga PN, Omondi D, Mireji PO, et al.: High-resolution melting analysis reveals low Plasmodium parasitaemia infections among microscopically negative febrile patients in western Kenya. Malar J. 2014; 13: 429. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChua KH, Lim SC, Ng CC, et al.: Development of High Resolution Melting Analysis for the Diagnosis of Human Malaria. Sci Rep. 2015; 5: 15671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZianni MR, Nikbakhtzadeh MR, Jackson BT, et al.: Rapid discrimination between Anopheles gambiae s.s. and Anopheles arabiensis by high-resolution melt (HRM) analysis. J Biomol Tech. 2013; 24(1): 1–7. PubMed Abstract | Free Full Text\n\nKang D, Sim C: Identification of Culex complex species using SNP markers based on high-resolution melting analysis. Mol Ecol Resour. 2013; 13(3): 369–376. PubMed Abstract | Publisher Full Text\n\nWinder LC, Phillips C, Richards N, et al.: Evaluation of DNA melting analysis as a tool for species identification. Methods Ecol Evol. 2011; 2(3): 312–320. Publisher Full Text\n\nDistefano G, Caruso M, La Malfa S, et al.: High resolution melting analysis is a more sensitive and effective alternative to gel-based platforms in analysis of SSR--an example in citrus. PLoS One. 2012; 7(8): e44202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWittwer CT: High-resolution DNA melting analysis: advancements and limitations. Hum Mutat. 2009; 30(6): 857–859. PubMed Abstract | Publisher Full Text\n\nTong Y, Jiang C, Yuan Y, et al.: Molecular identification of antelope horn by melting curve analysis. Mitochondrial DNA. 2014; 26: 1–7. PubMed Abstract | Publisher Full Text\n\nOsathanunkul M, Suwannapoom C, Osathanunkul K, et al.: Evaluation of DNA barcoding coupled high resolution melting for discrimination of closely related species in phytopharmaceuticals. Phytomedicine. 2016; 23(2): 156–165. PubMed Abstract | Publisher Full Text\n\nLi J, Song M, Xiong C, et al.: Application of barcode high-resolution melting for rapid authentication of the medicinal plant Psammosilene tunicoides. Biotechnol Biotechnol Equip. 2016; 30(4): 790–796. Publisher Full Text\n\nGanopoulos I, Bazakos C, Madesis P, et al.: Barcode DNA high-resolution melting (Bar-HRM) analysis as a novel close-tubed and accurate tool for olive oil forensic use. J Sci Food Agric. 2013; 93(9): 2281–2286. PubMed Abstract | Publisher Full Text\n\nSang R, Kioko E, Lutomiah J, et al.: Rift Valley fever virus epidemic in Kenya, 2006/2007: the entomologic investigations. Am J Trop Med Hyg. 2010; 83(2 Suppl): 28–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMontero-Pau J, Gómez A, Muñoz J: Application of an inexpensive and high-throughput genomic DNA extraction method for the molecular ecology of zooplanktonic diapausing eggs. Limnol Oceanogr Methods. 2008; 6: 218–222. Publisher Full Text\n\nKearse M, Moir R, Wilson A, et al.: Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012; 28(12): 1647–1649. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFolmer O, Black M, Hoeh W, et al.: DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol. 1994; 3(5): 294–299. PubMed Abstract\n\nZhang Z, Schwartz S, Wagner L, et al.: A greedy algorithm for aligning DNA sequences. J Comput Biol. 2000; 7(1–2): 203–214. PubMed Abstract | Publisher Full Text\n\nGarritano S, Gemignani F, Voegele C, et al.: Determining the effectiveness of High Resolution Melting analysis for SNP genotyping and mutation scanning at the TP53 locus. BMC Genet. 2009; 10: 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScott JA, Brogdon WG, Collins FH: Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg. 1993; 49(4): 520–529. PubMed Abstract\n\nTahir HM, Mehwish, Kanwal N, et al.: Genetic diversity in cytochrome c oxidase I gene of Anopheles mosquitoes. Mitochondrial DNA. 2015; 12: 1–4. PubMed Abstract | Publisher Full Text\n\nMcGee MD, Neches RY, Seehausen O: Evaluating genomic divergence and parallelism in replicate ecomorphs from young and old cichlid adaptive radiations. Mol Ecol. 2016; 25(1): 260–268. PubMed Abstract | Publisher Full Text\n\nHebert PD, Penton EH, Burns JM, et al.: Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proc Natl Acad Sci U S A. 2004; 101(41): 14812–14817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRuiz F, Linton YM, Ponsonby DJ, et al.: Molecular comparison of topotypic specimens confirms Anopheles (Nyssorhynchus) dunhami Causey (Diptera: Culicidae) in the Colombian Amazon. Mem Inst Oswaldo Cruz. 2010; 105(7): 899–903. PubMed Abstract | Free Full Text\n\nMurugan K, Vadivalagan C, Karthika P, et al.: DNA barcoding and molecular evolution of mosquito vectors of medical and veterinary importance. Parasitol Res. 2016; 115(1): 107–121. PubMed Abstract | Publisher Full Text\n\nLyman DF, Monteiro FA, Escalante AA, et al.: Mitochondrial DNA sequence variation among triatomine vectors of Chagas' disease. Am J Trop Med Hyg. 1999; 60(3): 377–386. PubMed Abstract\n\nAjamma YU, Mararo E, Omondi D, et al.: Dataset 1 in: Rapid and high throughput molecular identification of diverse mosquito species by high resolution melting analysis. F1000Research. 2016. Data Source"
}
|
[
{
"id": "15904",
"date": "12 Sep 2016",
"name": "Michael Zianni",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAjamma have an article on the use of the technique high resolution melt analysis towards the identification of morphologically indistinct species of mosquito. The specific goal is to expand the current set of primers in the research literature in order to identify more species from multiple genera of mosquitoes.\nThe methodology is clear with sufficient details for it to be reproduced by listing all appropriate reagents, DNA primer sequences and real-time PCR instrumentation. I do recommend adding the criteria by which the melt curves were deemed to be sufficiently different to allow identification of the species as compared to the \"Did Not Separate\" state as reported in Table 4.\nFigures 1 - 3 and 5 are clear and support the results summarized in table 4. I appreciate the authors efforts to repeat and report data from the previously published primers (\"AgamUni\") as a point of comparison. Appropriate controls were used with (1) water as a negative control for amplification and (2) samples from defined colonies and samples previously sequenced as positive controls. The most significant limitation is the number of replicates, and the diversity of sample collection points for each species.\n\nThe authors clearly acknowledge these limitations in the conclusion and clearly state the need for additional samples to asses the intra-specific variation which is critically important information to make this method highly useful.\nIn summary, the paper is clearly and concisely written with 1 minor recommendations for additional information on the method. The goals of the research are clearly stated, and the results as well as the conclusions support the goals. The researchers have achieved the goals by identifying and confirming at least one primer pair for each of 4 genera that identify various species that are difficult to identify by morphology alone.",
"responses": []
},
{
"id": "16385",
"date": "19 Sep 2016",
"name": "Hwa Chia Chai",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall, this is a well-written article on the development of HRM for identification of different species of mosquitoes. The authors also mentioned about the limitations of the study. However, the six pairs of primers used in this study did not seem to have clear directions or purpose. Did the authors want to come out with a pair of universal primers to differentiate all the clinically important mosquito species mentioned in the article? Or one specific primer pair for each genus of mosquitoes? Or there is other intention? Although the authors showed the ability of primer pairs to discern some mosquito species, for instance, COI-AnophF/HCO2198R could distinguish Mansonia spp., Aedes spp. and Culex spp., how about the species of the same genera that could not be discriminated? It would be more focused and directional if the primer pairs are genus-specific.\n\nMethods:\nIt would be great if the clinical importance of or disease transmitted by each mosquito genus/species is listed in Table 1 and Table 2 so that the readers understand the significance of discriminating the mosquito species.\n\nPlease mention the Tm for each primer in Table 3.\n\nPlease elaborate more on the primer design: on what basis the authors design the primers and for what reason they want to amplify those regions with amplicon sizes mentioned in Table 3.\n\nWhat was the amount of DNA template used for HRM analyses?\n\nResults:\nPlease show the limitation of detection for each assay in the detection of each mosquito species.\n\nWas auto-calling mode used for clustering? What was the confidence interval? Please mention the mean Tm for each species with standard deviation.\n\nFigure 2A, B, D: There was only one sample of C. tenagius included in the analysis. It is difficult to conclude the melting profile of this species if there was only one sample available.\n\nIn Table 1, COI-AnophF/HCO2198R has been shown to be able to separate Cx. tenagius from Cx. antennatus, Cx. pipiens from Culex sp. GPA, and Cx. neavei from Cx. duttoni. Since I am not an expert in mosquitoes, is there any specific reason the authors wanted to report the separation between two species of Culex rather than reporting it as a separation between all species?\n\nIn Figure 2A, the melt curves of Cx. antennatus, Cx. neavei, Culex sp. GPA and Cx. pipiens looks closely apart and hard to differentiate. Was the clustering auto- or manually called? What was the confidence interval if it was auto-called? What is the possibility of all these species being present or analysed at the same time? It would be hard to differentiate them if all of them are present in a same run of HRM analysis.\n\nHow do the melt curves of the unseparated species appear as compared to those in the same genus which could be differentiated? For instance, Cx. perexiguus vs. Cx. tenagius, Cx. antennatus, Cx. pipiens, Culex sp. GPA, Cx. neavei and Cx. duttoni. And also melting curves of Aedeomyia (Ad. africana and Ad. furfurea), Mimomyia (Mi. hispida and Mi. splendens) and Coquillettidia (Cq. aurites, Cq. chrysosoma, Cq. fuscopennata, Cq. metallica, Cq. microannulatus, Cq. pseudoconopas and Cq. versicolor) species (Table 4) or among An. funestus and An. coustani species complexes, which could not be differentiated in this study.\n\nIn Figure 4, intraspecies variation in the target sequence is seen, for instance, in the four strains of Ma. africana. Although the authors did mention on Page 6 “Although there are SNPs within species DNA that resulted to the slight changes observed in their HRM profiles, the SNPs across species were enough to distinguish between them”. I am not sure how confident it is to employ these assays in the presence of intraspecies variations, plus the sample size may be too small to validate the assays.\n\nFigure 5: please revise the figure legend.\n\nDiscussion:\nOn page 8, 2nd paragraph, the authors mentioned “Surprisingly, HRM analysis of PCR products from the COIAnophF/HCO2198R primers, which were designed for Anopheles, could not distinguish between these sibling species….” Can the authors explain the possible reason for this?\n\nAgain, since I am not a mosquito expert, are the mosquitoes of the same genus morphologically identical? If the assays are successfully developed, are the mosquitoes going to be morphologically identified prior to subjecting them to HRM analysis? How are the authors going to decide which primer pairs to use later in the mosquito identification?",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1949
|
https://f1000research.com/articles/5-1946/v1
|
10 Aug 16
|
{
"type": "Method Article",
"title": "Unit testing, model validation, and biological simulation",
"authors": [
"Gopal P. Sarma",
"Travis W. Jacobs",
"Mark D. Watts",
"S. Vahid Ghayoomie",
"Stephen D. Larson",
"Richard C. Gerkin",
"Gopal P. Sarma",
"Travis W. Jacobs",
"Mark D. Watts",
"S. Vahid Ghayoomie",
"Richard C. Gerkin"
],
"abstract": "The growth of the software industry has gone hand in hand with the development of tools and cultural practices for ensuring the reliability of complex pieces of software. These tools and practices are now acknowledged to be essential to the management of modern software. As computational models and methods have become increasingly common in the biological sciences, it is important to examine how these practices can accelerate biological software development and improve research quality. In this article, we give a focused case study of our experience with the practices of unit testing and test-driven development in OpenWorm, an open-science project aimed at modeling Caenorhabditis elegans. We identify and discuss the challenges of incorporating test-driven development into a heterogeneous, data-driven project, as well as the role of model validation tests, a category of tests unique to software which expresses scientific models.",
"keywords": [
"model",
"biology",
"software",
"validation",
"C. elegans",
"testing",
"quality control"
],
"content": "Introduction\n\nSoftware plays an increasingly prominent role in the biological sciences. This growth has been driven by an explosion in the availability of data and the parallel development of software to store, share, and analyze this data. In addition, simulations have also become a common tool in both fundamental and applied research1,2. Simulation management (initialization, execution, and output handling) relies entirely on software.\n\nSoftware used for collaborative biological research has an additional level of complexity (beyond that shared by other widely-used software) stemming from the need to incorporate and interact with the results of scientific research, in the form of large datasets or dynamical models. This added level of complexity suggests that technical tools and cultural practices for ensuring software reliability are of particular importance in the biological sciences3.\n\nIn this article, we discuss our experience in applying a number of basic practices of industrial software engineering—broadly known as unit testing and the related concept of test-driven development4–7—in the context of the OpenWorm project. OpenWorm (http://www.openworm.org) is an international, collaborative open-science project aimed at integrating the world’s collective scientific understanding of the C. elegans round worm into a single computational model8. It is a diverse project incorporating data, simulations, powerful but intuitive user interfaces, and visualization. Since the goal of the project is to simulate an entire organism, the project and its underlying code are necessarily complex. The scope of the project is immense – OpenWorm has over fifty contributors from sixteen countries and projects divided into over forty-five sub-repositories under version control containing a total of hundreds of thousands of lines of code. For a project of this magnitude to remain manageable and sustainable, a thorough testing framework and culture of test-driven development is essential4–7. In Figure 1, we show a diagrammatic overview of the many projects within OpenWorm and the relationship of testing to each of these. For extremely small projects, unit testing simply adds an overhead with little or no return on the time investment. As the project grows in size, however, the gains are quite significant, as the burden on the programmers of maintaining a large project can be substantially reduced.\n\nIn the code excerpts below, we will discuss 4 types of tests that are used in the OpenWorm code-base. They are:\n\nVerification tests: These are tests of basic software correctness and are not unique to the scientific nature of the project.\n\nData integrity tests: These are tests unique to a project which incorporates data. Among other purposes, these tests serve as basic sanity checks verifying, for instance, that each piece of data in the project is associated with a scientific paper and corresponding DOI.\n\nBiological integrity tests: These are tests that verify correspondence with known information about static parameters that characterize C. Elegans, for example, the total number of neurons.\n\nModel validation tests: These are tests unique to projects which incorporate dynamic models. Model validation tests (using the Python package SciUnit) verify that a given dynamic model (such as the behavior of an ion channel) generates output that is consistent with known behavior from experimental data.\n\nThe target audience for this article is computational biologists who have limited experience with large software projects and are looking to incorporate standard industrial practices into their work, or who anticipate involvement with larger projects in either academia or industry. We also hope that the exposition will be accessible to other scientists interested in learning about computational techniques and software engineering. We hope to contribute to raising the quality of biological software by describing some basic concepts of software engineering in the context of a practical research project.\n\n\nUnit testing for scientific software\n\nThe basic concept behind software testing is quite simple. Suppose we have a piece of code which takes some number of inputs and produces corresponding outputs. A unit test, verification test, or simply test is a function that compares an input-output pair and returns a boolean value True or False. A result of True indicates that the code is behaving as intended, and a result of False indicates that it is not, and consequently, that any program relying on that code cannot be trusted to behave as intended.\n\nLet us take a simple example. Suppose we have a function that takes a list of numbers and then returns them in sorted order, from lowest to highest. Sorting is a classic algorithmic task, and there are many different sorting algorithms with different performance characteristics; while the specific strategies they employ differ wildly, ultimately the result should be the same for any implementation. A unit test for one’s sorting algorithm should take as input a list of numbers, feed it to the sorting algorithm, and then check that each element in the output list is less than or equal to the one that comes after it. The unit test would return True if the output list had that property, and False if not.\n\nIf one has multiple implementations of a sorting algorithm, then one can use a reliable reference implementation as a testing mechanism for the others. In other words, a test might return True if a novel sorting algorithm gives the same result as one widely known to be valid. There are other strategies along these lines. For example, suppose we have an implementation of an algorithm for multiplication called multiply. If we have a trusted implementation of an algorithm for addition, we can test that our multiplication algorithm works as expected by checking its behavior against the appropriate number of addition operations, e.g., multiply(3,5) = 3 + 3 + 3 + 3 + 3. See Listing 1 for an implementation of this test in Python code.\n\nListing 1. Simple test for the multiplication operation.\n\n\n\nIn the previous example, the hypothetical unit test verified the core functionality of the algorithm. We had an algorithm that claimed to sort things, and we wanted to check that it worked as advertised. But there are many other kinds of tests that we might be compelled to write in order to know that our software is working correctly. For instance, what happens if we feed an empty list to our sorting algorithm (this is an example of an edge case)? Should it simply return the list, generate an error message, or both? What if a user accidentally gives the algorithm something that is not a list, say for example, an image? What should the error message be in this case? Should there be a single error message to cover all cases, or should the error message be tailored to the specific case at hand? One can easily write unit tests to verify that the correct behavior has been implemented in all of these cases.\n\nThe sum total of all of the desired behaviors of an algorithm is called a specification, or spec for short. For instance, the specification for a sorting algorithm might look like the following:\n\nWhen given a list of numbers, return the list sorted from smallest to largest.\n\nWhen given a list of strings, return the list sorted in lexicographic order.\n\nIf the input is an empty list, return the empty list and do not generate an error message.\n\nIf the input is not a list, generate the error message “Input should be a list of real numbers or strings”.\n\nIf the input is neither a list of strings nor a list of numbers, return the same error message as above.\n\nIn Listing 2, we have given a suite of unit tests for a sorting algorithm called mySort based on this specification. The basic notion demonstrated in the context of the sorting algorithm extends to any piece of software. In OpenWorm, we make extensive use of unit testing to verify both the functional properties of the system, as well as the validity of the data and models that comprise the simulation. For instance, the two tests given below in Listing 3 check that any worm model has 302 neurons, and that the number of synapses for a given type of neuron is in accordance with its known value from the scientific literature. We will examine the different types of tests in more detail in the next section.\n\nListing 2. Sample tests for the sorting specification given in the text. The class SortingTest is a container for all of the individual tests that define the specification and can be extended if more tests are added.\n\n\n\nListing 3. Excerpts from basic biological integrity tests for worm models. Given the size of the data repositories that OpenWorm relies upon, there are many simple tests such as these for ensuring the correctness of the associated data.\n\n\n\nIn test-driven development, the specification for a piece of software, as well as the corresponding unit tests are written before coding the software itself4,7. The argument for test-driven development is that having a well-developed testing framework before beginning the actual process of software development increases the likelihood that bugs will be caught as quickly as possible, and furthermore, that it helps the programmer to clarify their thought processes. In practice, while some tests are written before-hand, others are written in parallel with the rest of code development, or shortly after a piece of code is written but before it is integrated.\n\nWe mention here that, in the software community, a distinction is often made between unit tests and integration tests7. Strictly speaking, a unit test is a test which is applicable to the smallest, functional unit of code, and which has no external dependencies. On the other hand, tests which verify that different components work together are classified as integration tests; they verify that multiple components are integrated correctly. Some of the tests discussed below would strictly be considered integration tests. For the sake of simplicity, we will not distinguish between unit tests and integration tests in this article, and will refer to both as simply tests or unit tests. The primary distinction that we make here is instead between ordinary verification tests (to verify that code works as intended) and model validation tests (to validate a model against experimental data), which we discuss in more depth below.\n\nThe software that makes up OpenWorm shares common ground with all other pieces of software, whether the sorting algorithm described above, a word processor, or an operating system. As a result, there are a range of unit tests that need to be written to ensure that basic pieces of the software infrastructure function correctly. Many of these tests will not be of any scientific significance; they are simply sanity checks to ensure correct behavior for predictable cases. For instance, there are tests for checking that certain internal functions return the appropriate error messages when given incorrect inputs; there are tests for verifying that databases are loaded correctly; there are tests which check that functions adhere to a specific naming convention which will help automated tools process the code-base.\n\nAs a data-driven, scientific research project, however, OpenWorm also makes use of several other categories of tests that do not typically appear in software development. For instance, the PyOpenWorm subproject of OpenWorm is a simple API that provides a repository of information about C. elegans anatomy (https://github.com/openworm/PyOpenWorm). Given that the aim of OpenWorm is to produce a realistic simulation of the nematode, a carefully curated repository of empirical information is a cornerstone of the project.\n\nIn the context of unit testing, there needs to be a category of tests that ensure that a curated datum has been appropriately verified and, furthermore, that its internal representation in the PyOpenWorm database is consistent. For example, for each “fact” in PyOpenWorm, there needs to be an associated piece of evidence, which serves as a reference. Practically, this evidence consists of a Digital Object Identifier9, or DOI, which corresponds to a research paper from which the fact was originally retrieved. For this class of tests, we traverse the database of facts and verify that for each fact there is an associated source of evidence, i.e., a DOI. Furthermore, these tests verify that each DOI is valid, and that the URL corresponding to the DOI is accessible. There are also tests to check the internal consistency of the PyOpenWorm database, for instance, that neurons with the same name have the same identifier. Listing 4 gives several excerpts from the PyOpenWorm testing framework. It consists of tests to verify the references in the database, i.e., the DOIs which correspond to research papers.\n\nListing 4. Verifying data integrity is an integral component of testing in OpenWorm. Below, we give several sample tests to verify the existence of valid DOIs, one technique used to ensure that facts in the PyOpenWorm repository are appropriately linked to the research literature.\n\n\n\nIn Listing 5, we give several tests for verifying the contents of the PyOpenWorm repository. Since each of the functions below is designed to test properties of Neuron objects, they are part of a single class called NeuronTest. These tests fall into the category of verification tests, and several of the tests, such as test_name and test_type simply check that the database is working correctly.\n\nListing 5. An assortment of verification tests from PyOpenWorm. These verify that the database behaves as we would expect it to, that properties of certain objects (Neuron objects, in this case) are correctly specified, and that the database is not populated with duplicate entries.\n\n\n\nMany computational models in biology are compared only informally with the experimental data they aim to explain. In contrast, we formalize data-driven model validation in OpenWorm by incorporating tests to validate each dynamical model in the project against experimental data from the literature. As an example, consider a scenario where a developer creates a new model and provides parameter values for a simulation. In addition to running all of the verification tests described above, the model and parameter values must be validated with respect to established experimental results. In general, each summary output of the model is validated against a corresponding piece of data. One example of a summary model output is the “IV Curve” (i.e. current evoked in response to each of a series of voltage steps) of a given neuronal ion channel. We expect that our model will possess only ion channels which behave similarly to those observed experimentally, i.e. that the model IV Curve matches the experimentally-determined IV curve. If our model’s IV curve deviates too greatly from that observed experimentally, the model developers should be alerted and provided with information that will allow them to investigate the source of the discrepancy10. This may mean that parameter values must be modified, or in some cases the model itself must be substantially revised. In the case of OpenWorm, the necessary data for validating models is part of the PyOpenWorm and ChannelWorm subprojects (https://github.com/openworm/ChannelWorm), which are repositories of curated information about C. elegans anatomy and ion channels.\n\nOrdinary unit testing frameworks do not readily lend themselves to this kind of model validation. Rather than simply comparing an input-output pair, model validation tests should perform the same procedure that a scientist would perform before submitting a newly hypothesized model for publication. That is, they should generate some kind of summary statistic encoding the deviation between experimental data and model output. For example, in the case of an IV Curve, one might use the area between the model and data curves as a summary statistic. In the case of OpenWorm, because these models are part of a continuously updated and re-executed simulation, and not simply static equations in a research paper, the model validation process must happen automatically and continuously, alongside other unit tests.\n\nTo incorporate model validation tests, we use the Python package SciUnit11 (http://sciunit.scidash.org). While there are some practical differences between writing SciUnit tests and ordinary unit tests, the concepts are quite similar. For example, a SciUnit test can be configured to return True if the test passes, i.e. model output and data are in sufficient agreement, and False otherwise. Ultimately, a scientific model is just another piece of software—thus it can be validated with respect to a specification. In the case of dynamical models, these specifications come from the scientific literature, and are validated with the same types of tests used before submitting a model for publication. SciUnit simply formalizes this testing procedure in the context of a software development work-flow. In Listing 6, we give an example of SciUnit tests using the neuron-specific helper library NeuronUnit (http://neuronunit.scidash.org) for neuron-specific models.\n\nListing 6. Excerpt from a SciUnit test in ChannelWorm, a repository of information about ion channels. The test listed here verifies that a given ion channel has the correct current / voltage behavior. In terms of the informal classification of tests given above, this test falls under the category of model validation tests.\n\n\n\nIn the preceding example, the statistic is computed within the SciUnit method judge, which is analogous to the self.assert statements used in the ordinary unit tests above. While the ordinary unit test compares the output of a function pair to an accepted reference output, judge compares the output of a model (i.e. simulation data) to accepted reference experimental data. Internally, the judge method invokes other code (not shown) which encodes the test’s specification, i.e. what a model must do to pass the test. The output of the test is a numeric score. In order to include SciUnit tests alongside other unit tests in a testing suite, they can be configured to map that numeric score to a boolean value reflecting whether the model/data agreement returned by judge is within an acceptable range.\n\nThe output of these model validation tests can also be inspected visually; Figure 2 shows the graphical output of the test workflow in Listing 6, and illustrates for the developers why the test failed (mismatch between current-voltage relationship produced by the model and the one found in the experimental literature). Further details about the output of this test – including the algorithm for computing model/data agreement, and the magnitude of disagreement required to produce a failing score – can also be accessed via attributes and methods of the score object (not shown, but see SciUnit documentation). Consequently, full provenance information about the test is retained.\n\nSome computational science projects use ad-hoc scripts that directly run models and compare their outputs to reference data. This can be adequate in simple cases, but for larger projects, particularly distributed open-source projects with many contributors, the mixing of implementation and interface carries significant drawbacks12. For example, in order to record and store the membrane potential of a model cell–to then compare to reference data–one could determine which functions are needed to run the simulation in a given simulation engine, extract the membrane potential from the resulting files, and then call those functions in a test script. However, this approach has three major flaws. First, it may be difficult for a new contributor or collaborator to understand what is being tested, as the test code is polluted with implementation details of the model that are not universally understood. Second, such a test will not work on any model that does not have the same implementation details, and thus has limited re-usability. Third, any changes to the model implementation will require parallel changes to the corresponding tests. In contrast, by separating tests from implementation details, tests can work on any model that implements a well-defined set of capabilities exposed via an interface. SciUnit does this by design, and SciUnit tests interact with models only through an interface of standard methods, for example, those provided by NeuronUnit. It is the responsibility of the model developer to match this interface by referencing standard methods, e.g. run, get_membrane_potential, etc. Ultimately, the separation of implementation from interface leads to greater code clarity, more rapid development, and greater test re-usability.\n\nThe coverage of a testing suite is defined as the percentage of functions in a code-base which are being tested. Since there is no rigorous measure of what constitutes an adequate test, precise figures of test coverage should be interpreted with caution. Nonetheless, automated tools which analyze a code-base to determine test coverage can be a valuable resource in suggesting areas of a code-base in need of additional attention. Ideally, test coverage should be as high as possible, indicating that a large fraction of or even the entire code-base has been tested according to the intended specifications.\n\nIn PyOpenWorm, we make use several of pre-existing tools in the Python ecosystem for calculating test coverage of the Python code-base, specifically, the aptly-named Coverage package13, as well as a GitHub extension dedicated to tracking the coverage of such projects known as Coveralls14. We adopted these tools in an effort to track which parts of the code-base need additional tests, and to give further backing to the test-driven culture of the project. PyOpenWorm currently has a test coverage of roughly 73%. If a contributor to PyOpenWorm introduces some new code to the project but does not add tests for it, the contributor will see that test coverage has been reduced. By making changes in test coverage explicit, for example with a badge on the project’s homepage, it is easier to track the impact of a growing code-base.\n\nModern software is often written using a process of continuous integration or CI15,16, whereby the contributions of developers are integrated into a shared repository multiple times a day by an automated system. Typically, the output of a testing suite will determine whether or not the new contributions of a developer can be immediately integrated, or whether changes are required to avoid regression, i.e. failing unit tests that passed before the new contribution.\n\nThe benefits of continuous integration include early detection of bugs, eliminating development bottle-necks close to the release date (in the case of commercial software), and the regular availability of usable versions of the software. The process of continuous integration also encourages shifts in how developers think about structuring their code, and encourages regular, modular contributions, rather than massive, monolithic changes that can be difficult to debug.\n\nThe entire OpenWorm project, including the PyOpenWorm and ChannelWorm modules make use of continuous integration (see Figure 3), taking advantage of a free service called Travis-CI (https://travis-ci.org) that tests changes to the code-base as they are pushed to the collaborative software development portal GitHub17. With each change, the entire project is built from scratch on a machine in the cloud, and the entire test suite is run. A build that passes all tests is a “passing build”, and the changes introduced will not break any functionality that is being tested. Because the entire project is built from scratch with each change to the code-base, the dependencies required to achieve this build must be made explicit. This ensures that there is a clear roadmap to the installation of dependencies required to run the project successfully – no hidden assumptions about pre-existing libraries can be made.\n\nEach row corresponds to a single set of contributions, known as a commit, submitted by a given developer. A commit is assigned a build number, which is given in the second column, and the result of the build process is indicated by the color of the corresponding row. If any of the unit tests fail, the build will be marked as failed (errored, in red), and the code contributions will be rejected. The developer is then responsible for identifying and fixing the corresponding bugs, and resubmitting their contributions to the code repository.\n\nSuppose we have rigorously employed a process of test-driven development. Starting with a carefully designed specification, we have written a test suite for a broad range of functionality, and are using a continuous integration system to incorporate the ongoing contributions of developers on a regular basis.\n\nIn this scenario, given that we have written a test suite prior to the development of the software, our CI system will reject all of our initial contributions because most tests fail, simply because the code that would pass the tests has not been written yet! To address precisely this scenario, many testing frameworks allow tests to be annotated as expected failures or simply to skip a given test entirely. The ability to mark tests as expected failures allows developers to incrementally enable tests, and furthermore draws attention to missing functionality. Consequently, the fraction of tests passed becomes a benchmark for progress towards an explicit development goal, that goal being encoded by the set of all tests that have been written.\n\nThe OpenWorm code-base makes extensive use of skipped tests and expected failures as a core part of the culture of test-driven development. In PyOpenWorm, for example, data integrity tests are often added in advance of the data itself being incorporated to the database. These tests provide a critical safety net as new information is curated from the scientific literature. Prior to the curation of this information, the tests are simply skipped. Once the information is curated, the tests are run, and indicate whether the information is usable by the project.\n\nTests are typically sufficiently straightforward to write that it is easy to proliferate a testing suite with a large number of unnecessary tests. Often, these tests will be completely frivolous and cause no harm, beyond causing a testing suite to take much longer than necessary to run. However, tests which are overly specific can actually hinder the process of development. If there are tests which are too specific and constrain internal behavior that is not meant to be static, a developer’s contributions may be unnecessarily rejected during the process of continuous integration.\n\n\nConclusions\n\nOur aim in this article is to give an overview of some basic development practices from industrial software engineering that are of particular relevance to biological software. As a summary, we list here the types of tests used in OpenWorm. This list is simply an informal classification, and not a definitive taxonomy:\n\nVerification tests (the usual suspects) These are tests common to all pieces of software and are not particularly relevant to the biological nature of the project. For instance, tests that verify that error handling is implemented correctly, that databases are accessed correctly, or that performing certain numerical operations produces results within an acceptable range.\n\nData integrity tests These are tests unique to a project that incorporates curated data. In the case of OpenWorm, these tests check (among other things) that every biological fact in the PyOpenWorm repository has an associated piece of experimental evidence, typically corresponding to a DOI, and that each of these DOIs is valid.\n\nBiological integrity tests These tests verify that data tokens in the PyOpenWorm repository correspond to known information about C. Elegans. In contrast to the model validation tests described below, biological integrity tests typically only check static information/parameters.\n\nModel validation tests These are tests specific to a project that incorporates scientific models. Model validation tests allow us to check that specific models, such as the behavior of ion channels, correspond to known behavior from the scientific literature. In effect, they extend the notion of unit testing to compare summary data and model output according to some summary statistic. In OpenWorm, the Python package SciUnit and derivative packages like NeuronUnit are used for writing tests that check the validity of scientific models against accepted data.\n\nIt should be clear from the above discussion and corresponding code examples that unit tests are fundamentally quite simple objects. Their behavior is no more than to compare input-output pairs, or in the case of SciUnit tests, that a given model’s output corresponds to a known reference from the scientific literature. The sophistication of testing frameworks is generally quite minimal when compared to the software itself being tested. While ad-hoc test scripts may be sufficient for small projects, for large projects with many contributors, a systematic approach to unit testing can result in significant efficiency gains and ease the burden of long-term code maintenance. In the context of continuous integration, whereby a piece of software is built in an ongoing cycle as developers make changes and additions to the code-base, unit testing provides a valuable safety net that can prevent flawed code from prematurely being integrated.\n\nHowever, in spite of the conceptual simplicity and potential pitfalls of testing, its importance cannot be overstated. Writing tests requires careful thought and planning and some knowledge of the code-base being tested. Testing from a specification alone can result in inadequate testing, but tests which are too specific to the code-base can result in unnecessary roadblocks for developers.\n\nRather than being thought of as a sophisticated set of technical tools, unit testing should be viewed as a cultural practice for ensuring the reliability of complex software. Perhaps a useful analogy is the powerful impact that checklists have had in clinical medicine, aviation, construction, and many other industries18–20. Unit tests are sanity checks at a minimum, and can potentially guide the scientific development of models when used in conjunction with experimental data. In order to reap their benefit, their existence and maintenance needs to be valued by all of the participants of the research and software development process. Finally, in order for this culture to be created, test-driven development should not be a heavy-handed imposition on the developers. Otherwise, it will be incorrectly perceived as a bureaucratic hurdle, rather than the valuable safety-net that it is.\n\n\nSoftware availability\n\nSoftware available from: http://www.openworm.org/\n\nLatest source code: http://github.com/OpenWorm",
"appendix": "Author contributions\n\n\n\nGPS, TWJ, SDL, and RCG wrote the manuscript. All authors contributed to the unit testing framework in the Open Worm project.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded in part by NIMH (R01MH106674, RCG), and NIBIB (R01EB021711, RCG; and R01EB014640, Sharon M. Crook).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank supporters of OpenWorm, including NeuroLinx.org, the backers of the 2014 OpenWorm Kickstarter campaign (http://www.openworm.org/supporters.html), Google Summer of Code 2015, and the International Neuroinformatics Coordinating Facility. We would also like to thank the scientific and code contributors to OpenWorm (http://www.openworm.org/people.html), and Shreejoy Tripathy for careful reading of the manuscript.\n\n\nReferences\n\nTakahashi K, Yugi K, Hashimoto K, et al.: Computational Challenges in Cell Simulation: A Software Engineering Approach. IEEE Intelligent Systems. 2002; 17(5): 64–71. Publisher Full Text\n\nMacklin DN, Ruggero NA, Covert MW: The future of whole-cell modeling. Curr Opin Biotechnol. 2014; 28: 111–115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGewaltig MO, Cannon R: Current practice in software development for computational neuroscience and how to improve it. PLoS Comput Biol. 2014; 10(1): e1003376. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeck K: Test Driven Development: By Example. Addison Wesley, 2002. Reference Source\n\nMaximilien EM, Williams L: Assessing test-driven development at IBM. In Software Engineering, 2003. Proceedings. 25th International Conference on. IEEE, 2003; 564–569. Publisher Full Text\n\nErdogmus H, Morisio M, Torchiano M: On the effectiveness of the test-first approach to programming. IEEE Transactions on Software Engineering. 2005; 31(3): 226–237. Publisher Full Text\n\nOsherove R: The Art of Unit Testing: with examples in C#. Manning Publications, 2013. Reference Source\n\nSzigeti B, Gleeson P, Vella M, et al.: OpenWorm: an open-science approach to modeling Caenorhabditis elegans. Front Comput Neurosci. 2014; 8: 137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDOI® System and Internet Identifier Specifications. 2015. Accessed: 2015-07-24. Reference Source\n\nDe Schutter E: The dangers of plug-and-play simulation using shared models. Neuroinformatics. 2014; 12(2): 227–228. PubMed Abstract | Publisher Full Text\n\nOmar C, Aldrich J, Gerkin R: Collaborative infrastructure for test-driven scientific model validation. In Companion Proceedings of the 36th International Conference on Software Engineering. ACM Press, 2014; 524–527. Publisher Full Text\n\nShalloway A, Trott JR: Design patterns explained: a new perspective on object-oriented design. Pearson Education, 2004. Reference Source\n\nCode coverage measurement for python. 2015. Accessed: 2015-07-24. Reference Source\n\nCoveralls-Test Coverage History and Statistics. 2015. Accessed: 2015-07-24. Reference Source\n\nBooch G: Object Oriented Analysis and Design with Applications. Benjamin-Cummings, 1990. Reference Source\n\nDuvall PM, Matyas S, Glover A: Continuous Integration: Improving Software Quality and Reducing Risk. Addison-Wesley Professional, 2007. Reference Source\n\nTravis CI - Test and Deploy Your Code with Confidence. 2015. Accessed: 2015- 07-24. Reference Source\n\nGawande A: The Checklist Manifesto: How to Get Things Right. Picador. 2011. Reference Source\n\nHuang L, Kim R, Berry W: Creating a culture of safety by using checklists. AORN J. 2013; 97(3): 365–368. PubMed Abstract | Publisher Full Text\n\nWeiser TG, Berry WR: Perioperative checklist methodologies. Can J Anaesth. 2013; 60(2): 136–142. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15597",
"date": "24 Aug 2016",
"name": "Robert Cannon",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present an interesting review of how they have applied traditional software testing methodologies to the OpenWorm project. It provides a nicely balanced perspective on a subject that often leads to strong opinions.\nAs they stay, the objective is to provide a focused case study of how testing is applied in the OpenWorm project. As such, the title is somewhat generic: I would suggest adding OpenWorm in there and possibly mentioning that this is a case study.\n\nProbably the most novel part of the work is the incorporation of \"Model Validation Tests\" which serve to verify that the components, such as ion channel models, from which the model is built, behave in line with experimental data. The authors state that \"Ultimately, a scientific model is just another piece of software—thus it can be validated with respect to a specification.\" In a sense this is true, but, as Ref 10 points out1 the specification in the literature is often vague, incomplete or generally erroneous. The SciUnit \"judge\" method appears to be the answer to this, replacing the usual software testing \"assert\" function. Presumably a lot of the subtlety of the approach, and indeed the scientific input whether a model is indeed a good match to experiments, is embedded in the implementation of the various \"judge\" methods. Although it is not essential for this paper it would be interesting to see a little more of how this is done in the OpenWorm project.",
"responses": []
},
{
"id": "15594",
"date": "30 Aug 2016",
"name": "Christian Roessert",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article the authors show how industrial practices of unit testing and test-driven development can be used and extended for computational modelling in biological sciences. The manuscript is well written and provides clear examples making it easy to understand the basic concepts.\nI believe that establishing a culture of test-driven development in biological sciences is of great importance. However, in my view applying software engineering practices to computational modelling is often not as easy as depicted by the authors. I have the following suggestions to improve the manuscript:\n\nJudging the quality and validity of a computational model is a matter of scientific discussion and often cannot be easily reduced to a pass or fail decision in a model validation test. I would like to see a bit more detail on the transformation of the numeric score to a Boolean value used in the given ion channel test example but also on the general (statistical) concepts behind these decisions.\n\nTo iteratively improve a computational model, it is important to know not only if but also why a certain model fails or passes the model validation test. Since continuous integration systems are designed for simple verification tests: can the detailed results/figures and scores for each model validation test be shown directly on the CI dashboard? A discussion on the limits of current CI tools for biological modelling would be very helpful.\n\nWhile the calculation of ion channel dynamics for a model validation test is computationally relatively cheap, computations become much more expensive once full detailed cell models or even networks have to be computed to validate against e.g. in vivo recordings. In these cases, the testing framework becomes much more sophisticated than “simple objects” and free services like Travis-CI will likely not be able to provide the required computational power. Is there a certain limit for your model validation test concept you would consider in the OpenWorm project and in general? Are there any ideas how to overcome these limitations? A discussion on the limits of the presented framework would be appreciated.",
"responses": []
},
{
"id": "15595",
"date": "31 Aug 2016",
"name": "Andrew Davison",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article provides an introduction to automated software testing, its application to computational biology, and model validation as a form of testing, with examples taken from the OpenWorm project. The article is clearly written, and will be a helpful resource for computational biologists.\nThe article could be improved by a deeper discussion of some of the more difficult issues in the automation of model validation:\nwhat criteria to apply when transforming a numerical measure of closeness into a pass/fail? how to support the use of different criteria by different scientists, who might weigh the relative importance of particular validations very differently? how to handle contradictory experimental results?\nI would also be interested to read a discussion of possible improvements to continuous integration dashboards in the context of continuous validation, e.g. tracking the evolution of numerical validation results across model versions.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1946
|
https://f1000research.com/articles/5-1227/v1
|
06 Jun 16
|
{
"type": "Research Article",
"title": "Analyzing compound activity records and promiscuity degrees in light of publication statistics",
"authors": [
"Ye Hu",
"Jürgen Bajorath",
"Ye Hu"
],
"abstract": "For the generation of contemporary databases of bioactive compounds, activity information is usually extracted from the scientific literature. However, when activity data are analyzed, source publications are typically no longer taken into consideration. Therefore, compound activity data selected from ChEMBL were traced back to thousands of original publications, activity records including compound, assay, and target information were systematically generated, and their distributions across the literature were determined. In addition, publications were categorized on the basis of activity records. Furthermore, compound promiscuity, defined as the ability of small molecules to specifically interact with multiple target proteins, was analyzed in light of publication statistics, thus adding another layer of information to promiscuity assessment. It was shown that the degree of compound promiscuity was not influenced by increasing numbers of source publications. Rather, most non-promiscuous as well as promiscuous compounds, regardless of their degree of promiscuity, originated from single publications, which emerged as a characteristic feature of the medicinal chemistry literature.",
"keywords": [
"ChEMBL",
"publications",
"bioactivity",
"compound",
"promiscuity"
],
"content": "Introduction\n\nGiven the large volumes of compounds and activity data that are becoming available in the public domain1, mining of activity data can be expected to provide fresh insights into structure-activity relationships, compound distributions over current targets, or compound activity profiles. From activity data, target annotations of bioactive compounds can be systematically extracted and their current degree of promiscuity be determined2. In this context, one can distinguish between “good” and “bad” promiscuity; the latter resulting from assay artifacts due to, for example, undesired compound pan-assay interference3,4 or aggregator5 characteristics; the former from the ability of small molecules to specifically interact with multiple targets2. A reliable assessment of so-defined good promiscuity usually depends on high data confidence1,2. The ability to specifically engage in interactions with multiple targets provides the molecular basis of polypharmacology associated with drugs or other bioactive compounds6–8. Therefore, a quantitative assessment of promiscuity is helpful to estimate the magnitude of cross-reactivity within the current spectrum of active compounds and targets and establish networks of ligand-target interactions for the prioritization of promiscuous vs. selective candidate compounds. The universe of all possible ligand-target interactions will most likely never be fully explored and data incompleteness9 will -to a more or lesser extent- be omnipresent. However, currently accessible volumes of compound activity data are so large that we can expect to draw statistically meaningful trends from them, for example, in the study of structure-activity relationships and activity cliffs or compound activity profiles. Most recent analyses of compound promiscuity on the basis of high-confidence activity data from medicinal chemistry have revealed that compounds covering the current spectrum of thousands of targets are on average active against one or two targets10. This low degree of detectable promiscuity was found to be essentially stable over time, especially during periods of exponential compound data growth over the past decade11. Even the currently most extensively assayed compounds extracted from the PubChem BioAssay database12, tested against hundreds of targets, were on average only active against two or three targets13, which further supports a conservative of promiscuity among bioactive compounds. However, compound promiscuity was found to be consistently lower than promiscuity of approved drugs, with a mean of about four targets per drug14, again assessed on the basis of high-confidence activity data. These findings give rise to speculations concerning possible reasons for the higher degree of drug promiscuity13.\n\nIn our current study, we add an additional layer of information to the analysis of compound activity profiles and promiscuity by tracing activity annotations back to source publications and determining their distribution over the literature. Although elaborate databases such as ChEMBL15,16, the major public repository for compounds and activity data from medicinal chemistry, largely rely on the extraction of data from the literature, publication information has thus far not been taken into consideration when analyzing activity data on a large scale. Therefore, we have systematically generated compound activity records from original publications and also analyzed promiscuity in relation to publication statistics.\n\n\nMaterials and methods\n\nFrom the latest version of ChEMBL15,16 (release 21), compounds were assembled for which direct interactions (i.e. assay relationship type “D”) with single human protein targets at the highest confidence level (assay confidence score “9”) and defined potency measurements (Ki and/or IC50 values) were reported. All approximate measurements (e.g. “>”, “<”, or “~”) were disregarded. These compounds and their activity records were designated “set 1” and represented a high-confidence data set according to previously established confidence criteria17. For comparison, a “set 2” was collected consisting of compounds with defined potency values (excluding approximate measurements) for single human protein targets. Hence, in this case, no assay type and confidence criteria were applied. In both cases, only activity measurements were considered that were reported in original publications and all of these publication records were collected.\n\nCompound data sets 1 and 2 were further organized and analyzed on the basis of:\n\nPublications. Compounds and activity data were assigned to individual publications and grouped by publications using compounds, assays, and targets as criteria.\n\nActivity records. All individual compound-target combinations were determined to generate “activity records”. A compound might be tested against the same target in different assays reported in a single or multiple publications. In addition, potency values might vary across different assays and publications or might be referenced in other publications. Therefore, for each activity record representing a unique compound-target combination, all corresponding publications and potency values were collected and added to the record.\n\nCompounds. Publications and activity data were also grouped by compounds, leading to the definition of four subsets including compounds active against\n\n(A) a single target reported in a single publication;\n\n(B) a single target reported in more than five publications;\n\n(C) more than five targets reported in a single publication;\n\n(D) more than five targets reported in more than five publications.\n\nFor sets 1 and 2, the degree of promiscuity of a compound was defined as the number of targets it was reported to be active against2. Promiscuity degrees were determined and analyzed in light of publication statistics.\n\n\nResults and discussion\n\nGiven our data selection and curation criteria described above, set 1 contained 168,208 unique compounds that were tested in 31,578 assays against 1566 human targets, as reported in Table 1. These activity data were reported in 11,213 publications from 70 different medicinal chemistry journals. Table 2 lists the top-ranked journals where most of these publications appeared. These eight journals published ~97% of the qualifying papers. In addition, a total of 318,570 potency measurements were available and associated with 257,138 unique activity records, which were defined as individual compound-target entries containing all associated publications and qualifying potency measurements. In addition, set 2 comprised 293,736 compounds yielding 621,704 potency measurements against 2170 human targets (Table 1), which were reported in 19,528 publications from 90 journals (Table 1 and Table 2). A total of 471,442 unique activity records were obtained.\n\nFor sets 1 and 2, the number of compounds, assays, targets, activity records, and potency measurements is given. In addition, for both sets, the total number of publications and subsets reporting activity values from a single assay, multiple assays for the same target, or multiple assays for different targets are provided.\n\nThe top eight journals with more than 100 qualifying source publications for sets 1 and 2 are listed.\n\nTable 1 also reports the distribution of assays and targets over source publications. Of the nearly 11,000 papers associated with set 1, 4449 (~40%) and 1483 (~13%) reported activity data derived from a single assay and multiple assays for an individual target, respectively. The remaining ~47% of the publications reported activity from multiple assays for two or more targets. Similar observations were made for set 2 (Table 1). Publications were further organized with respect to increasing numbers of assays, targets, and active compounds (Figure 1). The majority of publications of sets 1 and 2 reported one or two assays for one or two targets, while ~9% (set 1) and ~14% (set 2) of the papers contained results for more than five assays. In addition, ~5% (set 1) and ~6% (set 2) of the publications reported activity data for more than five targets. On average, a set 1 and set 2 publication reported 2.8 and 3.4 assays for 2.2 and 2.4 targets and 16.7 and 17.3 active compounds, respectively (Figure 1). Hence, assay, compound, and target statistics were very similar for both sets.\n\nHistograms monitor the percentages of publications reporting increasing numbers of assays, targets, and compounds for (a) set 1 and (b) set 2, respectively. In addition, the mean and median values are provided.\n\nFrom set 1 and set 2 publications, a total of 257,138 and 471,442 unique activity records were extracted, respectively. These activity records were classified according to the number of publications from which they originated and the number of different potency values that were reported for each compound-target combination (Figure 2).\n\nActivity records from (a) set 1 and (b) set 2 are classified using a decision tree structure.\n\nFigure 2a shows that ~95% (244,775) of the set 1 activity records originated from a single publication. Most of these activity records (218,508) were associated with a single potency value. In addition, for 26,267 records, two or more potency values were available that mostly differed. Varying potency values typically resulted from different experiments. Of the 12,363 activity records originating from two or more publications, 7535 had varying potency values, whereas 4828 were associated with multiple instances of the same value, which was likely referenced from a previous publication. A similar distribution of activity records was observed for set 2 (Figure 2b). Taken together, the results revealed that more than 90% of all activity records resulted from a single publication most of which appeared between 2006 and 2014.\n\nSmall subsets of 328 (set 1) and 632 (set 2) activity records originated from more than 10 publications. Figure 3a (set 1) and Figure 3b (set 2) report the relationships between the number of publications and distinct potency values associated with these records. Up to 20 different potency values were frequently observed, which often spanned an unexpectedly large potency range of two or more orders of magnitude, as shown Figure 3c (set 1) and Figure 3d (set 2). Figure 4 shows exemplary compounds from such activity records, which further illustrate these findings. For example, the compound at the top was involved in two activity records with isoforms of carbonic anhydrase, a “classical” target, which were associated with 129 and 209 publications, respectively, appearing over a period of 12 years. In both instances, the range of 60 or 61 distinct potency values spanned nearly four orders of magnitude, revealing very large variations of experimental assessments.\n\nFor (a) 328 (set 1) and (b) 632 (set 2) activity records (compound-target combinations) originating from more than 10 publications, the number of publications is plotted vs. the number of different potency values that were reported. In addition, in (c) (set 1) and (d) (set 2), the percentages of activity records covering increasing logarithmic potency ranges are given, e.g. “> 3” refers to a potency range of more than three orders of magnitude.\n\nShown are four compounds from set 1, which were tested against a given target in many publications and for which the largest numbers of distinct potency values were reported. Publication and potency value statistics are provided. CHEMBLID gives the compound identifier in ChEMBL.\n\nFor each of 168,208 and 293,736 unique compounds from sets 1 and 2, the degree of promiscuity was determined, as reported in Table 3, revealing comparable distributions over degree intervals. Consistent with previous findings, the majority of bioactive compounds were found to interact with a single target10. The mean degree of promiscuity was 1.5 for set 1 and 1.6 for set 2 and the median degree was 1.0 in both cases, also consistent with earlier findings10. However, the low degree of promiscuity detected for set 2 was rather surprising because in this case, assay type and confidence criteria were not applied. The only requirement for set 2 compounds was the availability of clearly specified potency values for human protein targets, which resulted in promiscuity degrees very similar to set 1 having higher data confidence. These findings indicated that the requirement of explicit potency values alone limited the number of target annotations, although potency values for the same target often differed in their magnitude. Table 4 reports the publication frequency of all compounds in sets 1 and 2. Consistent with the results obtained for activity records, most of the compounds were only found in one publication, regardless of whether one or more targets were investigated.\n\nFor set 1 and set 2, the number (percentage) of compounds with increasing numbers of confirmed targets (degrees of promiscuity) is reported.\n\nFor set 1 and set 2, the number (percentage) of active compounds reported in increasing numbers of publications is given.\n\nPromiscuity was also assessed by directly focusing on source publications instead of activity records. The results are summarized in Figure 5. For both set 1 (Figure 5a) and set 2 (Figure 5b), target annotations of compounds across all promiscuity degrees mostly originated from a single publication, although multiple publications also contributed in many instances. There was no detectable correlation between promiscuity degrees and the number of source publications. Four subsets of compounds (A–D) were defined covering different ranges of promiscuity degrees and source publications. In set 1 (Figure 5a), 113,475 (67.5%; subset A) and 47 (0.03%; subset B) compounds with a promiscuity degree of 1 originated from a single and more than five publications, respectively. In addition, 1049 (0.6%; subset C) and 218 (0.1%; subset D) compounds with a promiscuity degree >5 originated from a single and more than five publications, respectively. Thus, activity data characterizing most of the highly promiscuous compounds were also reported in a single publication. Equivalent observations were made for compounds in set 2 (Figure 5b). The nine most promiscuous compounds from set 1 are shown in Figure 6. These compounds were annotated with 30 to 71 targets reported in one to more than 50 publications. Overall more than 86% of promiscuous compounds originated from single publications and there was no relationship between the degree of promiscuity and increasing numbers of source publications. Hence, current degrees of compound promiscuity could not be attributed to publication statistics and cumulative effects.\n\nIn (a) (set 1) and (b) (set 2), compounds with increasing numbers of targets (top to bottom) reported in increasing numbers of publications (left to right) are given in a matrix format. In addition, four compound subsets (A–D) are defined.\n\nShown are nine compounds displaying the largest degrees of promiscuity in set 1. Publication statistics are provided.\n\n\nConclusions\n\nIn this study, compound activity records were systematically extracted from original publications and their distribution was analyzed. Furthermore, publications were classified on the basis of activity records. For given compound-target combinations, potency value ranges from different experiments were often unexpectedly large, although only well-defined potency measurements were considered (Ki or IC50 values). At the same time, the exclusive consideration of numerically explicitly defined potency measurements for human targets led to essentially the same promiscuity estimates as the use of higher-confidence activity data taking assay type and confidence criteria into account. For promiscuity exploration on the basis of compound activity data, the immediate focus on source publications added an as of yet missing piece to the analysis puzzle. Since the majority of promiscuous compounds, regardless of their degree of promiscuity, were traced back to single publications, there was not notable bias due to publication frequency and statistics. Negative results are typically not reported in the scientific literature when known active compounds are re-tested on other potential targets. Therefore, test frequency does only influence publication frequency if positive results are obtained. Potential evidence for such effects is currently only available for very small numbers of active compounds, leading to an overall consistent picture of low promiscuity among bioactive compounds, consistent with earlier investigations.\n\n\nData availability\n\nThe data selection criteria specified herein make it possible to directly reproduce all data sets from ChEMBL version 21. However the data generated for this study are also made freely available on Zenodo: Compound activity records associated with original publications in ChEMBL 21, doi: 10.5281/zenodo.5168818.",
"appendix": "Author contributions\n\n\n\nJB conceived the study, YH planned and performed the analysis, YH and JB wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHu Y, Bajorath J: Learning from ‘big data’: compounds and targets. Drug Discov Today. 2014; 19(4): 357–360. PubMed Abstract | Publisher Full Text\n\nHu Y, Bajorath J: Compound promiscuity: what can we learn from current data? Drug Discov Today. 2013; 18(13–14): 644–650. PubMed Abstract | Publisher Full Text\n\nBaell JB, Holloway GA: New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J Med Chem. 2010; 53(7): 2719–2740. PubMed Abstract | Publisher Full Text\n\nBaell J, Walters MA: Chemistry: Chemical con artists foil drug discovery. Nature. 2014; 513(7519): 481–483. PubMed Abstract | Publisher Full Text\n\nMcGovern SL, Caselli E, Grigorieff N, et al.: A common mechanism underlying promiscuous inhibitors from virtual and high-throughput screening. J Med Chem. 2002; 45(8): 1712–1722. PubMed Abstract | Publisher Full Text\n\nPaolini GV, Shapland RH, van Hoorn WP, et al.: Global mapping of pharmacological space. Nat Biotechnol. 2006; 24(7): 805–815. PubMed Abstract | Publisher Full Text\n\nBoran AD, Iyengar R: Systems approaches to polypharmacology and drug discovery. Curr Opin Drug Discov Devel. 2010; 13(3): 297–309. PubMed Abstract | Free Full Text\n\nLu JJ, Pan W, Hu YJ, et al.: Multi-target drugs: the trend of drug research and development. PLoS One. 2012; 7(6): e40262. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMestres J, Gregori-Puigjané E, Valverde S, et al.: Data completeness--the Achilles heel of drug-target networks. Nat Biotechnol. 2008; 26(9): 983–984. PubMed Abstract | Publisher Full Text\n\nHu Y, Bajorath J: High-resolution view of compound promiscuity [version 1; referees: 3 approved]. F1000Res. 2013; 2: 144. PubMed Abstract | Publisher Full Text\n\nHu Y, Jasial S, Bajorath J: Promiscuity progression of bioactive compounds over time [version 1; referees: 2 approved, 1 approved with reservations]. F1000Res. 2015; 4(Chem Inf Sci): 118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Xiao J, Suzek TO, et al.: PubChem’s BioAssay database. Nucleic Acids Res. 2012; 40(Database issue): D400–D412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJasial S, Hu Y, Bajorath J: Determining the Degree of Promiscuity of Extensively Assayed Compounds. PLoS One. 2016; 11(4): e0153873. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu Y, Bajorath J: Monitoring drug promiscuity over time [version 2; referees: 3 approved]. F1000Res. 2014; 3: 218. PubMed Abstract | Publisher Full Text\n\nGaulton A, Bellis LJ, Bento AP, et al.: ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2012; 40(Database issue): D1100–D1107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBento AP, Gaulton A, Hersey A, et al.: The ChEMBL bioactivity database: an update. Nucleic Acids Res. 2014; 42(Database issue): D1083–D1090. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu Y, Bajorath J: Influence of search parameters and criteria on compound selection, promiscuity, and pan assay interference characteristics. J Chem Inf Model. 2014; 54(11): 3056–3066. PubMed Abstract | Publisher Full Text\n\nHu Y, Bajorath J: Compound activity records associated with original publications in ChEMBL 21. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "14194",
"date": "13 Jun 2016",
"name": "Kimito Funatsu",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript brings a great amount of investigative work, aiming to provide further insight regarding activity data and promiscuity degrees to publication statistics. The work was well conducted and the data was presented in a way that is rather informative for the readers. Besides some grammar mishaps and a few points that need clarification, I recommend this work to be accepted for indexation once the following considerations are addressed.\nThe relation between promiscuity, or activity data, and single publications is relevant and the conclusions states that well. I believe, however, that this work lacks mentioning the full impact of such discovery. Some discussion is presented, but certain aspects should be highlighted more. How does this new development fit with previous investigations? Why is this conclusion important for those working with activity data and promiscuity?\n\np.2 When presenting data organization, why did the author decide to group compounds based on 1 and 5 targets? Any particular reason for setting multiple targets as 5, and not 4, 6, etc.?\n\np.2 “a conservative of promiscuity”. I understand what do you mean by it, but it should be better phrased for clarity’s sake.",
"responses": []
},
{
"id": "14466",
"date": "20 Jun 2016",
"name": "Hans Matter",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis interesting contribution by Bajorath et al. extends previous database analysis work in order to identify and annotate promiscuous compounds. The authors extract activity information from public databases like ChEMBL and then trace back this information to the original primary scientific literature. It has been often documented that multiple compounds interact not only with single targets, but sometimes with many desirable and / or undesirable targets (off-targets). Further analysis of these polypharmacology findings is of great utility in understanding drug profiles and striving for the design of molecular with better overall profiles. Additional test campaigns after identification of bioactive compounds often reveal additional target-ligand interactions, both on undesirable ADMET targets (hERG, CYP, transporters) and selectivity off-targets (GPCRs, neighbouring proteins). However, these campaigns are expensive and will only systematically be conducted for molecules with interesting biological data and overall profile. Therefore for most compounds in the primary literature, only a single assay data point is reported to discuss the SAR of a particular series. It is very unlikely that this situation will significantly change in the near future.\nThe report title and abstract cover the content well. The chemoinformatics approach is well conducted, clearly described and can most likely be reproduced by others. The results are presented in a clear and interesting way and capture the interest of F1000Research readers. The large dataset for this analysis was made publically available. The authors might also want to mention, whether software tools and subroutines from their study are available. Therefore this contribution is an essential view on available data for polypharmacology studies and should be indexed in its present form.\nI suggest that chemical structures displayed in figures 4 and 6 should be annotated with their trivial naves or drug names, if available. Furthermore groupings of the targets for compounds in both figures by target families might be instructive to see, whether compounds like staurosporine or flavones have only been tested for kinases or in a much broader manner.\nFurthermore implications of these results should be clearly discussed in the paper. This could also prompt for additional suggestions and guidelines on conducting in-silico polypharmacology studies on these sparse data-matrices.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1227
|
https://f1000research.com/articles/3-53/v1
|
13 Feb 14
|
{
"type": "Web Tool",
"title": "wigExplorer, a BioJS component to visualise wig data",
"authors": [
"Anil S. Thanki",
"Rafael C. Jimenez",
"Gemy G. Kaithakottil",
"Manuel Corpas",
"Robert P. Davey",
"Rafael C. Jimenez",
"Gemy G. Kaithakottil",
"Manuel Corpas",
"Robert P. Davey"
],
"abstract": "Summary: wigExplorer is a BioJS component whose main purpose is to provide a platform for visualisation of wig-formatted data. Wig files are extensively used by genome browsers such as the UCSC Genome Browser. wigExplorer follows the BioJS standard specification, requiring a simple configuration and installation. wigExplorer provides an easy way to navigate the visible region of the canvas and allows interaction with other components via predefined events.Availability: http://github.com/biojs/biojs; http://dx.doi.org/10.5281/zenodo.7721",
"keywords": [
"Numerous web applications exist for visualisation of biological data. Data can be prepared for visualisation using a variety of formats",
"one of which is the widely used wiggle (wig) file. A wiggle file contains text that defines either a feature or a data track. The wiggle format was developed by the UCSC genome browser1 and then quickly adopted by other initiatives2",
"3. Web applications such as genome browsers rely heavily on JavaScript",
"a popular language for processing and rendering client-side information in a web browser. Despite their widespread use in bioinformatics",
"biological web applications are usually implemented with no standard reutilisation guidelines in mind",
"hence BioJS was developed4."
],
"content": "Introduction\n\nNumerous web applications exist for visualisation of biological data. Data can be prepared for visualisation using a variety of formats, one of which is the widely used wiggle (wig) file. A wiggle file contains text that defines either a feature or a data track. The wiggle format was developed by the UCSC genome browser1 and then quickly adopted by other initiatives2,3. Web applications such as genome browsers rely heavily on JavaScript, a popular language for processing and rendering client-side information in a web browser. Despite their widespread use in bioinformatics, biological web applications are usually implemented with no standard reutilisation guidelines in mind, hence BioJS was developed4.\n\nBioJS is an open source JavaScript library of components for the visualisation of biological data on the web. Here we present wigExplorer, a standard, portable BioJS component designed to easily render wig data format files. wigExplorer can be integrated and controlled from other applications. To our knowledge, this is the first modular component to visualise wig data that complies with BioJS standards.\n\n\nThe wigExplorer component\n\nwigExplorer is fully integrated in the BioJS project. It follows the standards set by the BioJS registry5, a centralised repository of BioJS components hosted at the European Bioinformatics Institute (EBI). Having wigExplorer in the BioJS registry is advantageous because it promotes i) easy discoverability for the component, ii) collaborative development with other members of the BioJS community and iii) reutilisation by third party applications. In the BioJS registry, component APIs are exposed, i.e., events and methods are defined and documented so that other BioJS components can interact with each other. By following these conventions, wigExplorer is able to interact with other components on the same web page, enriching the overall experience for the user. The code below shows how to incorporate wigExplorer into a web application. Only three configuration elements are needed: the target HTML element in which the component will be rendered, the background colour of the component, and the file path containing the wig data.\n\n\n\nwigExplorer uses D3.js, the data-driven documents JavaScript library6, to generate graphical representations from wig data. D3.js handles the manipulation of the data documents, the reading of wig data as text format and their conversion to an area chart format. To control the visual aspect of the wig data, wigExplorer contains simple controls for zooming and panning.\n\nwigExplorer can be used to visualise genomic data in different ways. An application is shown in Figure 1, depicting single nucleotide polymorphism (SNP) density data from a genomic annotation in the tomato genome. Here chromosome 2 is zoomed in to show the genomic interval contained between position 3M and 47M. The SNP density data contained in the wig data file are presented as bins, where the Y axis indicates the number of SNPs contained in each bin. The screenshot shows a dramatic change in the density of SNPs at the 24M bin mark of the chromosome, suggesting a potential boundary for an introgression segment introduced from a closely related tomato species. Other potential applications of wigExplorer may involve the visualisation of gene expression and alignment data.\n\nThe top controls are designed to scroll sideways. Peaks show SNP density of 1kb size bins. A change of SNP density can be observed around the 29M mark, with a slightly greater density of SNPs on the right, indicative of a potential introgression segment from another related species.\n\nWe are aware that third party browsers are also using wigExplorer. A screenshot of the TGAC Browser7 is shown in Figure 2 using wigExplorer to depict Myzus spp. scaffold 1 zoomed in between regions 714K and 727K. Here strand-specific RNA-Seq paired-end read coverage is shown as a wig track. The track below shows a closely related annotated species gene set for comparison. This comparison suggests a potential gene extension in both forward and reverse orientation.\n\nThe wigExplorer track shows read coverage in Myzus spp. for scaffold 1. Forward and backward strands are depicted in red and blue respectively. Evidence genes from a closely related species are displayed in the track below.\n\n\nConclusions\n\nThe wigExplorer component provides a platform to visualise biological data in wig format. wigExplorer can be easily integrated with other web components or extended to provide new functionality. We expect this component to be particularly useful for visualisation in a variety of data types such as SNP density, alignments and gene expression. Like any other BioJS component, wigExplorer requires little technical knowledge for its utilisation.\n\n\nSoftware availability\n\nZenodo: wigExplorer, a BioJS component to visualise wig data, doi: 10.5281/zenodo.77218\n\nGitHub: BioJS, http://github.com/biojs/biojs;",
"appendix": "Author contributions\n\n\n\nAT and RJ developed the code for wigExplorer. MC and GK created the user cases for Figure 1 and Figure 2 respectively. AT, RD, MC and GK wrote the paper.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAT, GK, MC and RD were supported by the UK Biotechnology and Biological Sciences Research Council (BBSRC) National Capability Grant (BB/J010375/1) at TGAC.\n\n\nAcknowledgements\n\nWe are grateful to all BioJS developers who have contributed their work under an open source license. We are grateful to David Swarbreck at TGAC for his advice on the data shown in Figure 2.\n\n\nReferences\n\nKent WJ, Sugnet CW, Furey TS, et al.: The human genome browser at UCSC. Genome Res. 2002; 12(6): 996–1006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDonlin MJ: Using the Generic Genome Browser (GBrowse). Curr Protoc Bioinformatics. John Wiley and Sons, Inc., 2009. PubMed Abstract | Publisher Full Text\n\nSkinner ME, Uzilov AV, Stein LD, et al.: JBrowse: A next-generation genome browser. Genome Res. 2009; 19(9): 1630–1638. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGómez J, García LJ, Salazar GA, et al.: BioJS: an open source JavaScript framework for biological data visualization. Bioinformatics. 2013; 29(8): 1103–1104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBioJS: registry. http://www.ebi.ac.uk/Tools/biojs/registry/, 2013. Reference Source\n\nD3.js data-driven documents. http://d3js.org, 2012. Reference Source\n\nThanki AS, Bian X, Davey RP, et al.: TGAC Browser: visualisation solutions for big data in the genomic era. 2013. Reference Source\n\nThanki AS, Jiminez RC, Kaithakottil GK, et al.: wigExplorer, a BioJS component to visualise wig data. Zenodo. 2014. Data Source"
}
|
[
{
"id": "3699",
"date": "25 Feb 2014",
"name": "Robert Buels",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript is well written, but it does not accurately describe the software in question.I installed the software on my machine, and evaluated its behavior under several scenarios that I think would be commonly encountered by people using it. The major issues I found with the software are:wigExplorer does not provide any useful error messages when it fails to fetch, or fails to parse, the file specified in its `dataSet` option. This can be a major issue for prospective users (i.e. installers) of the software.Also, when I installed it and ran it via the included TestwigExplorer2.html, the \"top controls ... designed to scroll sideways\" described in the manuscript were not shown. To rule out any problems with the way I installed it (caused by missing stylesheets or the like), I inspected the browser's (Google Chrome 32) document object model (DOM) using its built-in development tools, and verified that the controls were not even being placed in the DOM by the code itself. I didn't debug any further.Most importantly, and I think this is a show stopper: wigExplorer does not actually support the Wiggle format as set forth by UCSC at http://genome.ucsc.edu/goldenPath/help/wiggle.html. It supports a limited tab-separated value format, which bears some resemblance to BedGraph format, but is not BedGraph either. So: it is named for a standard data format that it does not support, and the data format that it actually supports is not a standard format.Unfortunately, I cannot approve this manuscript as it is currently written. It does not accurately describe the wigExplorer software.",
"responses": [
{
"c_id": "719",
"date": "26 Feb 2014",
"name": "Anil Thanki",
"role": "Reader Comment",
"response": "Hi Robert,I thank you for your useful comments, and I fully understand your concerns about the code and the performance of the tool. I have already started working on fixes regarding your comments, and happy to say that the first two of your three comments have already been fixed. I am now working on the third point and hope to have a solution as soon as possible. Once i do have one I will submit a revised version of the new code and manuscript.Thanks a lot again,Anil"
},
{
"c_id": "841",
"date": "28 May 2014",
"name": "Anil Thanki",
"role": "Author Response",
"response": "Dear Robert We would like to thank you for taking the time to review the manuscript. Please find below our responses to your comments. 1) wigExplorer does not provide any useful error messages when it fails to fetch, or fails to parse, the file specified in its `dataSet` option. This can be a major issue for prospective users (i.e. installers) of the software.Answer: After considering this comment we have added more informative error messages, depending on the errors that can occur while running the component, e.g. “Unknown format detected”, “No data for selected reference”, “File not found”, etc. This will help make any installation problems clearer to users during installation.2) Also, when I installed it and ran it via the included TestwigExplorer2.html, the \"top controls ... designed to scroll sideways\" described in the manuscript were not shown. To rule out any problems with the way I installed it (caused by missing stylesheets or the like), I inspected the browser's (Google Chrome 32) document object model (DOM) using its built-in development tools, and verified that the controls were not even being placed in the DOM by the code itself. I didn't debug any further.Answer: Outside of the BioJS registry environment, the component’s CSS behaviour changes, which resulted in disappearing controls. We have reformatted the CSS for this component and this is now fixed.3) Most importantly, and I think this is a showstopper: wigExplorer does not actually support the Wiggle format as set forth by UCSC at http://genome.ucsc.edu/goldenPath/help/wiggle.html. It supports a limited tab-separated value format, which bears some resemblance to BedGraph format, but is not BedGraph either. So: it is named for a standard data format that it does not support, and the data format that it actually supports is not a standard format.Answer: We thanks Robert t raise this concern and we have rewritten the code and now the component fully supports the standard fixedStep and variableStep wig file format (as set out by UCSC) with span information and multiple references in one file if available."
}
]
},
{
"id": "3697",
"date": "07 Mar 2014",
"name": "Phil Lord",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI have tried this piece of software and it does appear to do what it says and running the software was not too difficult. The link to the software on Zenodo in the Software Availability section (Zenodo: wigExplorer, a BioJS component to visualise wig data, doi:10.5281/zenodo.77218) seems a bit pointless as the download bundle requires a large number of dependencies which are not included and for which paths have been hardcoded tolocal machine locations. The authors should also include either a tag or a commit hash so that the version of the software in the download bundle can be identified in GitHub.I ran the software from a GitHub checkout - here all the paths given in the documentation assume a specific configuration of directories (which is different from that of the GitHub repository). None of this is a show stopper, but it does make me rather doubt the validity of the statement: \"Like any other BioJS component, wigExplorer requires little technical knowledge for its utilisation.\"The article is, essentially, an advert with relatively little technical content. This is appropriate; there is little point including, for example, precise instructions for the use of wigExplorer as this article will be checkpointed and the code will become out of date. I would however have liked a bit more information on the expected performance of the software: for instance, how big a chromosome can it visualize? How manipulable is the visualisation?The writing in the article is okay, but there are a few parts which could do with improvement. For example, in the first paragraph of the Introduction section, the statement \"biological web applications are usually implemented with no standard reutilisation guidelines in mind, hence BioJS was developed\" I am not at all sure what \"reutilisation guidelines\" means with respect to code base. Do the authors just mean a defined and documented API? In which case, they should probably say this.There are also a couple of places which could do with simplification. In the second paragraph of the Introduction section:\"To our knowledge, this is the first modular component to visualise wig datathat complies with BioJS standards.\" This statement could be seen as a caveat, and I think the authors could risk saying \"This is the only BioJS component to show wig data\". Also in the second paragraph of the Application section the statement \"We are aware that third party browsers are also using wigExplorer.\" could perhaps be simplified to \"Third party browsers are also using wigExplorer\".",
"responses": [
{
"c_id": "832",
"date": "23 May 2014",
"name": "Anil Thanki",
"role": "Author Response",
"response": "We would like to thank Phil for taking the time to review the manuscript. Please find below our responses to your comments1) The link to the software on Zenodo in the Software Availability section (Zenodo: wigExplorer, a BioJS component to visualise wig data, doi:10.5281/zenodo.77218) seems a bit pointless as the download bundle requires a large number of dependencies which are not included and for which paths have been hardcoded to local machine locations. The authors should also include either a tag or a commit hash so that the version of the software in the download bundle can be identified in GitHub.Answer: We have uploaded a new version of the code to Zenodo with dependencies provided (http://dx.doi.org/10.5281/zenodo.8516). This new version is also included in BioJS Release 1.0 https://github.com/biojs/biojs/releases/tag/v1.02) It does make me rather doubt the validity of the statement: \"Like any other BioJS component, wigExplorer requires little technical knowledge for its utilisation.\" Answer: We understand that there are a number of required dependencies for this component, and not including these in the component’s distribution made it harder to install and utilise. As above, we have included these dependencies as well as sample datasets, so simply unzipping the distribution is sufficient to install and start using the component.3) The article is, essentially, an advert with relatively little technical content. This is appropriate; there is little point including, for example, precise instructions for the use of wigExplorer as this article will be checkpointed and the code will become out of date. I would however have liked a bit more information on the expected performance of the software: for instance, how big a chromosome can it visualize? How manipulable is the visualisation?Answer: F1000Research requires a working copy of the code when article get published (for archival purposes) so, we have uploaded current version of code to zenodo (https://zenodo.org/record/8516#.U334LlhdV0s) so this copy of code will be same though main code can be updated. So this article will always point to the version of the code that it describes, which is stored in zenodo. And on the performance issue, in this case size of the file matters rather than size of genomic region. Because size of data depends on depth of the genomic information, though we tried to make an effort to make it clear in manuscript. In first paragraph of wigExplorer component section, “Wig files contains minimalistic information of genomic data, wigExplorer can handle large genomic region like a chromosome (tested with a single file containing 12 chromosome with average length of 60 Mb), but it depends on depth of data rather than length of genomic region.”4) The writing in the article is okay, but there are a few parts, which could do with improvement. For example, in the first paragraph of the Introduction section, the statement \"biological web applications are usually implemented with no standard reutilisation guidelines in mind, hence BioJS was developed\" I am not at all sure what \"reutilisation guidelines\" means with respect to code base. Do the authors just mean a defined and documented API? In which case, they should probably say this. Answer: Here by using the word \"reutilisation guidelines”, we mean that BioJS components can be easily implemented and can interact with each other with standard API. We have added a statement explaining this in the last line of the first paragraph of introduction that “BioJS code contains standard guidelines that how to use components and how API can be implemented to interact with other components.”5) There are also a couple of places, which could do with simplification. In the second paragraph of the Introduction section:\" To our knowledge, this is the first modular component to visualise wig data that complies with BioJS standards.\" This statement could be seen as a caveat, and I think the authors could risk saying \"This is the only BioJS component to show wig data\". Also in the second paragraph of the Application section the statement \"We are aware that third party browsers are also using wigExplorer.\" could perhaps be simplified to \"Third party browsers are also using wigExplorer\". I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Answer: We have reformatted the first statement to “Currently, this is the only BioJS component to show wig data” and we have changed the second statement as reviewer suggested, “Third party browsers are also using wigExplorer”"
}
]
}
] | 1
|
https://f1000research.com/articles/3-53
|
https://f1000research.com/articles/5-1938/v1
|
09 Aug 16
|
{
"type": "Research Article",
"title": "Perceptions of hospital medical personnel on disaster preparedness",
"authors": [
"Maciej Walczyszyn",
"Shalin Patel",
"Maly Oron",
"Bushra Mina",
"Maciej Walczyszyn",
"Maly Oron",
"Bushra Mina"
],
"abstract": "Objective: Natural disasters, domestic terrorism and other forms of catastrophe, though rare, pose a significant public health challenge when they do occur. Hospital personnel must have the appropriate training to identify, treat, and possibly even oversee local disaster preparedness initiatives. Insufficient resources have been placed on the education received by healthcare providers in tertiary medical institutions. We intended to assess the current state of knowledge and interest in disaster preparedness among different tiers of hospital staff and training levels in order to identify potential barriers and areas for further training.\n\nDesign: A cross-sectional online survey was given to hospital attending physicians, subspecialty fellows, residents, nurses, physician assistants, and their respective students. The survey questions were disseminated throughout the Society of Critical Care Medicine (SCCM) Members and the North Shore Long Island Jewish (NSLIJ) hospital system via e-mail newsletters.\n\nMain results: A total of 572 individuals participated between October 2013 and May 2014. 85% of respondents expected to be dealing with a disaster during their career. 61.5% of respondents noted they would not feel comfortable leading and directing a local disaster management initiative. Yet 51.9% of respondents treated victims of natural disasters, 56.5% of transportation disasters and 34.8% of a structural collapse. When asked about level of formal disaster management training: 27.5% noted that no training was provided and 33% noted that they received 12 hours of training and only a quarter had more than 48 hours of formal training. 86.6% of respondents noted an interest in participating in a disaster management training workshop.\n\nConclusions:\nMany of our respondents had low level of disaster management training, did not feel comfortable leading a disaster initiative, however many have had to take care of victims of disasters. Based on our findings, hospital professionals feel under prepared for disaster management, and disaster preparedness should be considered an integral part of medical training.",
"keywords": [
"Disaster management",
"ebola virus"
],
"content": "Introduction\n\n“There's no harm in hoping for the best as long as you're prepared for the worst.” – Stephen King.\n\nThe World Health Organization defines a disaster as “A serious disruption of the functioning of a community or a society causing widespread human, material, economic or environmental losses which exceed the ability of the affected community or society to cope using its own resources”1.\n\nTerrorist attacks, natural catastrophes, infectious epidemics, and other forms of disasters, though rare, pose a significant public health challenge when they do occur. Healthcare providers are the receiving end of casualties from a disaster in the community, and they must ensure the necessary training to lead disaster preparedness initiatives in the scenario that one does occur.\n\nInstances such as the 2001 New York City September 11th terrorist attacks, the 2005 New Orleans Hurricane Katrina, the 2010 Haiti Earthquake, the 2011 Tohoku Earthquake and Tsunami, the 2012 Hurricane Sandy, and more recently the 2014 Ebola Virus outbreak, all reveal that mass casualties do not enter the hospital all at once. Rather, most of these victims entered over a protracted period of time ranging from acute traumas within hours of the event to symptoms of post-traumatic stress disorder presenting months to years after. This can tremendously exhaust understaffed and undertrained hospital personnel.\n\nDisaster management initiatives have more often emphasized pre-hospital protocols and personnel preparation while insufficient resources have been placed on the education and training of the healthcare providers in tertiary medical institutions that receive disaster victims. This has been previously termed “ambulances to nowhere”2,3. Disaster training is rarely incorporated in neither undergraduate nor graduate medical education.\n\nWe intend to assess the current state of knowledge and interest in disaster preparedness among different tiers of hospital staff and training levels in order to identify potential barriers and areas for further training.\n\n\nMaterials and Methods\n\nA cross-sectional online survey was given to hospital attending physicians, subspecialty fellows, residents, nurses, physician assistants, respiratory therapists and their respective students. The survey questions were disseminated using a cloud based company, Monkey Survey, throughout the Society of Critical Care Medicine (SCCM) Members and the North Shore Long Island Jewish (NSLIJ) hospital system e-mail newsletters in October 2013. Participants were given an explanation of the intentions of the survey, which included agreement to the publication of the data. All project expenses were funded by Lenox Hill Hospital, a part of the NSLIJ health system.\n\nThere is no standardized test for preparedness. The survey questions were designed to assess the current level of medical training of the participants in their respective fields and asked about their perception of disasters occurring in their healthcare system (Dataset 1). Specifically, participants were asked if they had to deal with a disaster in the past or thought they would have to deal with a disaster in the future and which disaster they thought would be likely to occur. The survey also assessed for the participants’ current level of disaster management training, in what form they had received it, and if they would feel comfortable being involved in a disaster management scenario. The survey then further evaluated if the participants would like additional training for disaster management and gauged what type of training they would find most effective. Finally the survey assessed for any barriers to achieving this training.\n\nThe responses to the survey were electronically collected from October 2013 to May 2014. The results were generated in percentages and analyzed by the authors of the study and the Monkey Survey Company.\n\n\nResults\n\nA total of 572 individuals participated between October 2013 and May 2014. Over 83% of respondents were not NSLIJ employees and over 60% were physicians, of which 83% identified themselves as attending physicians. 62% of attending physicians identified themselves as critical care physicians. The remainder of participants consisted of 79 nurses (two of which were students), 25 physician assistants, and eight respiratory therapists (one of which was a student) (Table 1). Greater than 90% of respondents identified their current or planned future practice locale as urban or suburban (Figure 1).\n\nResponders were able to select more than one location of work.\n\nA vast majority of participants had managed victims of disaster situations in the past. Just over half of participants (52%) stated they had treated victims of natural disasters; 57% had treated victims of transportation disasters; 35% - of structural collapses; 28% - of industrial catastrophes; 15% - of terrorist attacks and 16% had treated victims of warfare (Figure 2). When asked of future expectations, 85% of respondents expected to deal with a disaster during their career, choosing natural disasters as the most likely expected culprit (3.69 on a scale of 0–5, with 5 being most likely). This was followed by industrial catastrophes at 3.16 and terrorist attacks at 2.66 (Figure 3a and 3b). When considering terrorist threats, most participants believed explosives (2.87 on a scale of 0–5, with 5 being most likely) were most likely to be the cause of harm in their areas, followed by biological weapons (2.39), chemical weapons (2.35) and nuclear radiation (2.15).\n\n(a) Types of disasters expected to be encountered on a scale of 1 (least likely) to 5 (most likely). (b) Do hospital personnel expect to treat patients of disasters?\n\nWhen asked about level of formal disaster management training, 28% of participants noted they received no training, 33% noted they received 12 hours of training or less, 10% had a training of at least 24 hours, 5% noted up to 48 hours of training, and 25% had more than 48 hours of formal training. Of those who had received training, 41% were offered lectures and hands-on scenario exercises, 34% attended a separate disaster management seminar, 30% felt that part of their training came from real life experience, 21% had had individual study, and for 13%, the training was part of a graduate curriculum. When asked where this training was offered, 35% of respondents stated they were offered a separate training course, 6% said that training was part of a residency program, 6% said it was part of a fellowship program, 4% were trained at a graduate school and 19% stated that training was offered via other methods.\n\nOf the surveyed participants, only 38% felt comfortable leading and directing a local disaster management initiative; however nearly all participants (90%) felt they would be able to participate in a disaster management scenario. A large majority of respondents (87%) expressed their interest in participating in a disaster management-training workshop. Of these, 78% were interested in learning focused ultrasound exams, 92% wanted to learn procedures that may be needed during a disaster and 92% wished to participate in simulation training (Figure 4). The major identified barrier to training was lack of time (80% of respondents), followed by availability of resources (63%), access to experts (45%), obtaining scenario exercises (36%) and lack of interest (22%). The preferred methods of training were via live lectures with accompanied scenario exercises (66%), on-line courses (24%) and live lectures only (3%); 6% of participants were not interested in a training workshop.\n\n\nDiscussion\n\nDisaster medical training of hospital personnel is known to be inadequate and prior disasters have highlighted this issue4. Most of our respondents worked in critical care settings, over a quarter had no disaster management training and most of them did not feel comfortable leading a disaster initiative; however, many have had to take care of victims of disasters, with greater than 85% of respondents expecting to deal with a disaster during their career. Despite time being the number one barrier to further training, the overwhelming majority of participants (87%) noted an interest in participating in a disaster management-training workshop. Most of our respondents would like to receive further training in the form of live lectures and scenarios with the use of ultrasound machines, common procedures and simulations.\n\nOf note, availability of resources and access to experts were both identified as barriers to training, partially due to lack of awareness of available resources.\n\nDisasters cannot always be predicted, nonetheless, they can and need to be prepared for. This preparation can likely be addressed with adequate funding and allocation of time during formal training of all relevant professions. Although not ideal, there are currently online resources and courses available, free of charge, as listed in the “Compendium of Disaster Health Courses” drafted by the National Center for Disaster Medicine and Public Health (https://ncdmph.usuhs.edu/Documents/NCDMPH_Compendium_V1.pdf). Hands-on training in the form of drills and simulation seem to be the way forward for preparedness; however, these are not yet readily available. The Canadian Forces Medical Service have training rotations involving all levels of hospital personnel, including administrators a form of training that dates back over 100 years and had helped prepare for World War One5. In 2002, the Society of Critical Care Medicine (SCCM) set up a program called Fundamentals of Disaster Management (FDM), a one-day course directed to healthcare professionals to treat victims of mass casualty events. Such training seems almost crucial for preparedness with disasters becoming more frequently encountered by healthcare providers.\n\n“Chance favors the prepared mind.” – Louis Pasteur\n\nWith regard to Ebola preparedness, Governor Andrew M. Cuomo of New York State had designated eight hospitals statewide to treat patients with Ebola. Protocols for identifying, evaluating and isolating patients who require care were created and sent to all hospitals, diagnostic and treatment centers and ambulance services. The Port Authority ensured that proper training was in place for all airport personnel, as well as ensuring deployment of two ambulances at each airport, aimed to safely transport potential patients with Ebola. In addition, the Metropolitan Transport Authority (MTA) worked to make sure that their employees had necessary equipment and training to protect themselves. Personnel from the Centers for Disease Control and prevention (CDC), Customs and Border Protection, and the US Public Health Service, had practice drills with scenarios in dealing with passengers who may have been infected with the virus at John F. Kennedy International airport in New York. There were screening questionnaires for passengers from West African nations6. These measures indicate preparedness for Ebola have been taken seriously to both pre-hospital and hospital levels.\n\n\nConclusions\n\nDisaster preparedness integrates a number of elements. In the recent cases of Ebola, for example, these include airport and airline personnel, transport services, emergency services and hospital personnel. At the hospital level, our survey suggests that staff are unprepared for a disaster and are not comfortable leading a disaster initiative, yet they are interested in further training. The lack of availability of training remains a large deterrent. Based on our survey results, we recommend that incorporating lectures, accompanied by scenario-based disaster preparedness should be considered as an integral part of medical training.\n\n\nData availability\n\nF1000Research: Dataset 1. Assessing perceptions of disaster preparedness survey, 10.5256/f1000research.8738.d1302347",
"appendix": "Author contributions\n\n\n\nMaciej Walczyszyn MD – Survey design and data collection.\n\nShalin Patel MD – Data collection, data analysis and writing the manuscript.\n\nMaly Oron MD – Writing the manuscript.\n\nBushra Mina MD – Research mentor.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by the Critical Care Department of Lenox Hill Hospital.\n\n\nReferences\n\nWorld Health Organization: Humanitarian Health Action, Definitions: emergencies. Reference Source\n\nFarmer JC, Carlton PK Jr: Providing critical care during a disaster: the interface between disaster response agencies and hospitals. Crit Care Med. 2006; 34(3 Suppl): S56–9. PubMed Abstract | Publisher Full Text\n\nBarbera JA, Macintyre AG, DeAtley CA: Ambulances to Nowhere: America’s Critical Shortfall in Medical Preparedness for Catastrophic Terrorism. BCSIA Discussion Paper 2001–15. Executive Session on Domestic Preparedness (ESDP) Discussion Paper 2001– 07. Boston, John F. Kennedy School of Government, Harvard University, October, 2001. Reference Source\n\nKing MA, Dorfman MV, Einav S, et al.: Evacuation of Intensive Care Units During Disaster: Learning From the Hurricane Sandy Experience. Disaster Med Public Health Prep. 2016; 10(1): 20–7. PubMed Abstract | Publisher Full Text\n\nMcAlister VC: Drills and exercises: The way to disaster preparedness. Can J Surg. 2011; 54(1): 7–8. PubMed Abstract | Free Full Text\n\nGovernor Cuomo Outlines Ebola Preparedness Plan For New York State. 2014. Reference Source\n\nWalczyszyn M, Patel S, Oron M, et al.: Dataset 1 in: Perceptions of Hospital Medical Personnel on Disaster Preparedness. F1000Research. 2016. Data Source"
}
|
[
{
"id": "16489",
"date": "03 Oct 2016",
"name": "Amir Khorram-Manesh",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article aims to report the current state of knowledge and interest in disaster preparedness among different tiers of hospital staff and training levels in order to identify potential barriers and areas for further training.\n\nThere is nothing new about the conclusion. \"At the hospital level staff are unprepared for a disaster and are not comfortable leading a disaster initiative, yet they are interested in further training\". This has been reported earlier and does not add anything new to the global knowledge. What is new in your report? Have you compared your data with other hospitals?\n\nWhich are the potential barriers for further training? Point them out and discuss.\n\nDifferent tiers of hospital staff cannot be limited to attending physicians, subspecialty fellows, residents, nurses, physician assistants, respiratory therapists and their respective students. Why and how were these participants chosen?\n\nDisaster preparedness knowledge consists of many factors such as knowledge about risk and vulnerability analysis, organizational belonging, how to initiate a disaster plan etc. Have you investigated all components included in disaster preparedness? Simply asked how many have read the disaster plan?\n\nTraining is one way to standardize the multidisciplinary management of and preparedness for a disaster or a major incident. The authors write about different type of training. What kinds of training are these? Are these evaluated? Do we know how much and how long training is needed? What is acceptable preparedness? I am not sure whether you talk about individual training or multidisciplinary training? Training on a \"patient\" or mass casualty training.\n\nWhen using surveys, you may end up with some problems. What are the pros and cons with your method?\n\nOne of the conclusions is that incorporating lectures, accompanied by scenario-based disaster preparedness, should be considered as an integral part of medical training. In which perspective and why? How about other healthcare categories? Do you mean medical training at medical school or at the hospital?\n\nThe lack of availability of training remains a large deterrent. What is lacking; time for training, economic support for training, or a validated training model?\n\nI have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.",
"responses": []
},
{
"id": "18175",
"date": "02 Dec 2016",
"name": "Jonathon R. Gray",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for the opportunity to review this article. We want to compliment the authors on tackling such an important subject. Overall we approve this article, but we have some comments that the authors could address.\n\nGeneral comments The article could be improved by broadening the context for the reader and potentially adding to the survey questions.\n\nTo broaden the readers understanding of context, we would have liked to see a more comprehensive literature review. We would also suggest that the researchers could interview or send another questionnaire to hospital managers to explore the issues from both sides. Training is important, but perhaps more important is a culture of regular practice. This will ensure that, for example, materials are in the right place, staff know where they are and how to use them. Linked to this we would be interested in whether staff have assigned roles in any emergency. Both these related points could have been explored in the survey with simple questions about response planning, practice and roles assigned to respondents?\n\nSpecific points to address Grammatical issues:\n\n“Infectious epidemics”: “epidemics” is sufficient, as by definition an epidemic is caused by an infectious agent\n\n“choosing natural disasters as the most likely expected culprit…” This doesn’t read well.\n\nIn Materials & Methods, “Monkey survey” should be “survey monkey”.\n\nStudy context: A brief description of hospital size would be valuable. Is it a general hospital? What population/area does it serve?\n\nSpecificity of language: In Results the authors’ say, “572 individuals participated”. We are not clear if that means 572 responded to the survey? This should be clear because sometimes to hide a low response rate people will mention the number that were invited to respond as “participants”. We suspect 572 were responses, but it would be helpful if that was made clear.\n\nThe authors mention the types of doctors and nurses that responded but not where they work. The speciality/department could be very relevant.\n\n“Over 83%...”: ”over” and a percentage is OK for a summary but in full text the authors should provide the exact figure. What does “over” mean? It could be anything between 83 and 100%.\n\n“Greater than 90% respondents…” Authors should provide the exact percentage.",
"responses": []
},
{
"id": "17798",
"date": "01 Aug 2017",
"name": "Pier Luigi Ingrassia",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThanks for submitting this paper which attempts to assess the current state of knowledge and interest in disaster preparedness among different tiers of hospital staff and training levels in order to identify potential barriers and areas for further training. Hospital disaster preparedness is one of the pillars of a good health system. The topic is therefore of relevant interest. Nevertheless, the level of novelty is quite questionable since literature offers a variety of other published researches with similar results.\nThe manuscript is easy to read, comprehensive and well managed. Clearly, lots of work has been done by all the authors. Nevertheless, many improvements are warranted.\n\nAbstract The abstract provides a concise summary of the content of the article. It describes adequately the objective, design, results and conclusions of the study. In the objective a more concise statement about the aim of the work would be more appropriate for the scope of the paragraph.\n\nIntroduction Although the rationale of the study and the significance of the problem are well explained, more emphasis about disaster preparedness at hospital level and the lack of personnel training would be very appreciated for the benefit of such study.\nIn the statement “… they must ensure the necessary training to lead disaster preparedness initiatives in the scenario that one does occur” it is not clear who might ensure the necessary training. A deeper description and a citation would strengthen the concept.\n\nMaterials and Methods The method are well described but it is unclear how the survey instrument was designed and whether it was validated. In addition, a description of the questionnaire should be added to the manuscript. Being the study grounded on survey tool the authors would pay more attention to these aspects. A major intervention is therefore strongly required.\nThe authors invited participants throughout the Society of Critical Care Medicine (SCCM) Member list and the North Shore Long Island Jewish (NSLIJ) hospital system e-mail newsletters in October 2013. It would be helpful to quantify the entire population, whether possible, in order to figure out the response rate.\nAuthors collected the level of trainers of respondents but results are not reported and not used for deeper analysis (i.e. differences in perception amongst different levels of training). Authors should also justify why respondents were asked about which disaster they thought would be likely to occur. It seems not in line with the study objective.\n\nResults The results are consistent but sometime redundant with the figures. Figure 2, for example, reports all the figures described in subparagraph Experience and perception making the reading not fluent and more difficult. Authors are also invited to include in this section precise data not approximations.\nFigure 1 reports demographic data that can be included in the table 1.\n\nDiscussion This sections results quite weak.\nThe authors report a list of training initiatives delivered by well known Institutions and Agencies. The authors are invited to better explain the meaning of the findings and why they are important, considering all possible explanations for the study results.\nAuthors should also relate study findings to those of other studies. Literature offers a variety of other published researches with similar results. Please consider the Mortelmans LJ et al. 20161 and Lim GH et al. 20132 as examples. In addition, the authors mention the “ambulance to nowhere” case. Questions raised by this case may have served as the motivation for authors’ study and deeper considerations could be presented.\nStudy limitations were not addressed by the authors. Authors are asked to elicited them.\n\nConclusion The Conclusions is a bit blurry. In our opinion authors have demonstrated that many of the respondents had low level of disaster management training and feel underprepared for disaster management. The lack of availability of training is reports to be the main deterrent.\nThe Ebola case should be removed from this sections since it is not a conclusion derived from the study findings.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1938
|
https://f1000research.com/articles/5-1932/v1
|
08 Aug 16
|
{
"type": "Research Article",
"title": "Cluster analysis of Plasmodium RNA-seq time-course data identifies stage-specific co-regulated biological processes and regulatory elements",
"authors": [
"Efejiro Ashano",
"Itunuoluwa Isewon",
"Jelili Oyelade",
"Ezekiel Adebiyi",
"Itunuoluwa Isewon",
"Jelili Oyelade",
"Ezekiel Adebiyi"
],
"abstract": "In this study, we interpreted RNA-seq time-course data of three developmental stages of Plasmodium species by clustering genes based on similarities in their expression profile without prior knowledge of the gene function. Functional enrichment of clusters of upregulated genes at specific time-points reveals potential targetable biological processes with information on their timings. We identified common consensus sequences that these clusters shared as potential points of coordinated transcriptional control. Five cluster groups showed upregulated profile patterns of biological interest. This included two clusters from the Intraerythrocytic Developmental Cycle (cluster 4 = 16 genes, and cluster 9 = 32 genes), one from the sexual development stage (cluster 2 = 851 genes), and two from the gamete-fertilization stage in the mosquito host (cluster 4 = 153 genes, and cluster 9 = 258 genes). The IDC expressed the least numbers of genes with only 1448 genes showing any significant activity of the 5020 genes (~29%) in the experiment. Gene ontology (GO) enrichment analysis of these clusters revealed a total of 671 uncharacterized genes implicated in 14 biological processes and components associated with these stages, some of which are currently being investigated as drug targets in on-going research. Five putative transcription regulatory binding motifs shared by members of each cluster were also identified, one of which was also identified in a previous study by separate researchers. Our study shows stage-specific genes and biological processes that may be important in antimalarial drug research efforts. In addition, timed-coordinated control of separate processes may explain the paucity of factors in parasites.",
"keywords": [
"RNA-seq",
"Plasmodium",
"regulatory elements",
"expression profiles",
"malaria"
],
"content": "Introduction\n\nIn the past two decades, there has been an extraordinary commitment to the control and elimination of malaria which has resulted to a significant decrease in global malaria morbidity and mortality (Rao, 2015). In the bid to finding new ideas to deal with the disease, we have witnessed a shift from orthodox independent studies in biochemistry and microbiology to using a more multidisciplinary and robust approach that embraces other fields such as computer science and mathematics. The so called “omics” science has been employed with some measure of success in malaria research (Sarker et al., 2013). Studies in the transcriptome of malaria causative parasites in both human and animal models have revealed the integral role of transcriptional regulation in Plasmodium biology. Taking advantage of the availability of the full genome sequence data of P. falciparum, the parasite's complex expression patterns through each developmental stage within its respective unique environment have been studied using microarray. More recently, whole transcriptome short-gun techniques (RNA-seq) have been shown to be a more accurate predictor of expression trends than micro arrays (Wang et al., 2009).\n\nFinding cis-regulatory elements using in-silico methods assumes that genes that share similar expression trends (i.e. that are “co-transcribed”) are likely to be controlled by a common regulatory element (Young et al., 2008). These potential regulatory elements with promoter functions can be found upstream of the expressed genes and will appear as conserved sequence motifs common to genes found in the same cluster, but scarce in the remainder of the genome (Young et al., 2008). Clustering methods, often used in the study of gene expression profiles, have also been applied to the analysis of time-course data for over a decade (Eisen et al., 1998; Lukashin & Fuchs, 2001). Clustering algorithms group gene expression profiles on the basis of a distance metric. Backed by the power of statistics, this approach has been used effectively as a tool for visualization of micro-array, and more recently, RNA-seq data in identifying groups of co-regulated genes (Nueda et al., 2014; Tarca et al., 2006).\n\nPresently, there is a very little information of the basal transcription machinery of Plasmodium species. Very few transcription factors have been identified and this is largely attributable to the AT-rich nature of the parasites genome which makes it difficult to identify regulatory elements within (Callebaut et al., 2005; Young et al., 2008). There might be an underlying reason for the paucity of transcription factors however (Subramaniam et al., 2001).\n\nBased on these motivations, we sought to analyze Plasmodium RNA-seq data to yield insight that could potentially be used in antimalarial drug discovery research. Our efforts primarily focused on identifying biological processes of therapeutic interest and implicated cis-regulatory elements involved in the coordinated regulation of these processes.\n\n\nMaterials and methods\n\nThe data used in this study, made freely available by Otto and colleagues (Otto et al., 2014), comprised 5020 normalized gene fragments per kilobase of exon per million (FKPM) expression values of Plasmodium berghei ANKA. Two replicates of gene expression values were collected at six time points, these time points corresponding to the ring form (RI and RI-R), the trophozoite (Tr and Tr-R), schizont (Sch, Sch-R), gametocyte (G and G-R), 16-hour ookinete (O-16, O-16R) and the 24 hour ookinete (O-24h) morphological phases of the parasite life cycle. These morphological phases span three life-cycle stages commonly named the intra-erythrocytic development cycle, gametocytes/gamete stage, and the fertilization and ookinete development (the last stage which occurs within the vector host). For convenience, these stages are indicated as IDC, SEX and MOS. The IDC time-points included the ring form, trophozoite and schizont time-points; the SEX included the ring form, trophozoite and gametocyte time-points; and the MOS included the gametocyte, 16-hour ookinete and the 24-hour ookinete time-points respectively.\n\nIdentification of statistically significant expressed (or repressed) genes (p < 0.05) and expression profile clustering for the IDC, SEX and MOS development stages of P. berghei was done in R (version 3.2.2) on a Windows 64-bit platform using the maSigPro package. maSigPro was initially designed for microarray time course experiments but has since been upgraded to handle next-generation sequencing (NGS) series data properly. It finds genes with significant temporal expression changes using a two-step regression strategy. For single time course experiment, the procedure first adjusts the global model by the least-squared technique to identify differentially expressed genes selecting significant genes with a false discovery rate (FDR) control procedure. The regression fit for each gene is defined by computing p-values associated to the F-Statistic of the model, which is used to select significant genes. P-values are corrected for multiple comparisons by applying the linear step-up (B-H) FDR procedure (Benjamini & Hochberg, 1995). Finally, a stepwise regression is used to find statistically significant different profiles. Significantly expressed genes with similar expression patterns are then clustered using a hierarchical clustering approach applying the coefficients obtained in this second regression model. (Nueda et al., 2014). The cut-off value for the R-squared of the regression value used for this study was 0.7. The number of clusters was set as 9.\n\nGene ontology (GO) analysis was performed with PlasmoDB (http://plasmodb.org/plasmo/) which sources GO information from Interpro (http://www.ebi.ac.uk/interpro/) and the Annotation Center. The top ranked enriched GO terms for biological processes were generally reported except in circumstances where no biological process was enriched with the gene set, in which case, GO terms for components were highlighted. A p-value cutoff was set to 0.05 (see supplementary file S01).\n\nThe Suite for Computational identification Of Promoter Elements, “SCOPE” motif finder (http://genie.dartmouth.edu/scope/) was used to predict candidate promoter motifs in this study. The SCOPE motif finder is designed to identify candidate regulatory DNA motifs from sets of genes that are coordinately regulated using three program algorithms (Carlson et al., 2007). A fixed upstream length of 1000 bp was maintained as default for Plasmodium species (Harris et al., 2011; Jurgelenaite et al., 2009). Input genes used for the detection of motifs included P. falciparum (3D7) orthologues of the corresponding P. berghei genes extracted from PlasmoDB. Consensus sequences were generally reported if they had a coverage of greater than 90%, a greater than 2-fold ratio of motif count to gene number per cluster, and a significant (sig) value (Carlson et al., 2007) that was greater than 10. A sig value of 0 implies that one motif of that significance is expected by chance. Carlson et al. (2007) demonstrated that sig values performance on synthetic data significantly improved after the value of 10 and remained fairly unchanged below the sig value of 55.\n\n\nResults and discussion\n\nMaSigPro identifies differentially expressed genes from FPKM normalized gene expression and selects genes that are significantly expressed applying false discovery rate control procedures. The data show that relatively fewer genes are up-regulated in the IDC (Table 1). Only 1448 genes of the 5020 genes (~29%) captured in the time-course experiment showed any significant activity. Furthermore, between the three clusters that were up-regulated in this stage, only 155 genes out of the 1448 (~0.1%) have been functionally annotated or identified. Likewise, data from earlier studies that used different investigatory approaches seem to support this observation. Early experiments by McGarvey et al. (1984) in recombinant clones corresponding to genes expressed specifically during the late schizont-merozoite stage of P. falciparum development showed that the maturation of the parasite in this stage was associated with the selective activation of a relatively small set of genes. In another study, the level and nature of transcriptional activity in P. falciparum and its role in controlling gene expression during the IDC was investigated using nuclear run-on on whole-transcriptome analysis (Sims et al., 2009). In this experiment, it was observed that the total transcriptional activity involved in the IDC was seen to peak late in the stage at the advent of other morphological stages (i.e. the SEX and MOS stage). A possible explanation to this was given in Bozdech et al. (2003) who suggested that Plasmodium species induce genes only when required and just once per cycle. Other possible reasons for this controlled regulation may be as an adaptive response by the parasite to evade the advanced immunological mechanisms of its human host and leaving the parasite with fewer targets for drugs.\n\nGenes can be clustered based on common function or sub-cellular complex membership, a process that has enjoyed some success over time. However, the functions of a large amount of the genes in Plasmodium are unknown, making this approach unsuited for a full analysis of the parasite’s genome especially for parasite-specific processes and sub-cellular structures. In addition, as seen in kinases and proteases involved in signal transduction, two genes of similar function may not share a common regulatory element, which further flaws this approach. On the other hand, steady state mRNA levels provide a more direct estimation of the transcriptional effects of cis-acting regulatory elements with genes of similar expression trends a better basis for identifying putative regulatory elements (Young et al., 2008; Zhou et al., 2005). Young et al. (2008) used a semi-supervised clustering algorithm OPI that utilized information on Plasmodium gene function sourced from the GO database to guide clustering of genes based on microarray derived life cycle mRNA expression patterns. In this study however, we opted for an unsupervised hierarchical clustering algorithm that altogether eliminated the bias brought into the clustering process with the prior knowledge from gene ontology (Herrero et al., 2001). This resulted in grouping highly co-expressed genes in clusters based simply on similarities in calculated expression patterns independent of gene function or sub-cellular complex membership. Results of identified associated GO biological processes and components enriched in clusters of their respective developmental stages with reference to studies of these processes in malaria control are shown in Table 2. The importance of these processes in parasite development and the relevance in malaria control are discussed in more detail in the following paragraphs. It is worth noting that in a previous microarray study on the transcriptome of the intraerythrocytic development cycle of Plasmodium falciparum, it was observed that induction of genes occurred selectively at specific times and only when required (Bozdech et al., 2003). Our results not only further support this report, but also suggest that the same applies for other stages of the Plasmodium development cycle.\n\nReferences of relevant processes targeted by on-going drug related research are shown in column 4.\n\nIt takes about 16 hours for a merozoite (via the ring form) to be transformed to a matured trophozoite just before it divides its nucleus. It takes another 6 – 8 hours for the schizogony to be complete unleashing a fresh set of meroziotes for reinvasion. Of the 5020 genes analyzed, a total of six clusters where identified to show up-regulated trends of fold changes greater than 2 from their median expression profiles in the three life stages of the parasite. For the IDC, two clusters showed significant up-regulated expression. These included cluster 4 (16 genes) which was up-regulated after the 16-hour time point; and cluster 9 (32 genes) which was up-regulated from the 0-hour time point up until the 24-hour time point (Figure 1). GO analysis of these clusters showed the former to be associated with genes involved in the “apical part of the cell” (GO:0045177), while the latter being associated with the “lipoate biosynthetic process” (GO:0009107). Only 5% of the genes compared to the background of genes found in the apical part of the cells were identified in cluster 4 so it is unlikely that the apical part of the parasite plays a significant role in in this development stage. Some of the genes identified within cluster 4 may however be important therapeutic targets. These include the TREP (TRAP-like protein) (PBANKA_1306500), SIAP1 (PBANKA_1006200) and PLP4 (PBANKA_0711400) proteins among other highly conserved proteins which have not at the time of writing this report been characterized (see supplementary file S02). SIAP1 and TRAP have been implicated in the invasion of sporozoites by facilitating binding to host cells. Knock-out experimentation of TRAP in P. berghei led to non-motile sporozoites and by such, has been thought to be an important therapeutic target (Ejigiri et al., 2012; Greenwood et al., 2008; Sultan et al., 1997). Other studies also confirm the expression of this protein during the IDC (Baum et al., 2008). A list of highly conserved coding genes with unknown function that can be studied for function and as potential drug targets in this and other clusters can be found in the supplementary file S02. Our results also show that Lipoate-protein ligase (LipB), one of the two genes involved in Lipoic acid synthesis is steadily increased throughout the IDC period. Lipoic acid is an integral cofactor of α-keto acid dehydrogenase complexes and the glycine cleavage system, the metabolite playing a dual role in both intermediate metabolism and as a redox sensor/antioxidant. Biosynthesis of lipoate occurs exclusively in the apicoplast involving LipB and lipoic acid synthase. There is evidence however that disruption of the LipB gene does not negatively affect the growth of the parasite suggesting the protein’s redundancy. Lipoate scavenging is another route for lipoate acquisition which is critical for Plasmodium survival especially in the liver stages of development and may be a better target than the actual lipoate synthesis by the parasite for drugs (Allary et al., 2007; Günther et al., 2009; Storm & Muller, 2012).\n\nIn each IDC a few parasites loop out of asexual multiplication and differentiate into sexual cells, otherwise known as gametocytes. These haploid macro-gametocytes (females) and micro-gametocytes (males) are the precursor cells of the female and male gametes. Some evidence shows that the parasite commits to forming gametocytes 12 – 16 hours after invasion. The factors that cause trophozoites to differentiate into gametocytes in preference to schizonts are not known. It is known however, unlike in P. falciparum, they do not have periods of “pure” differentiation where “all” trophozoites differentiate into gametocytes. It should be of note that gametocyte formation is necessary for the transmission of P. berghei since this is the only form by which the mosquito can take up the parasite during a blood meal (Landau & Gautret, 1998). In the SEX, two clusters showed regulatory patterns of interest. The median expression profile of cluster 2 (851 genes) was increased between the 0 hour and the 16 hour time points. This increase is seen to level off for the remaining part of the life-cycle stage. Cluster 5 (27 genes) also demonstrated an up-regulated expression profile, but only after the 16 hour time point (Figure 1). GO analysis of cluster 2 in SEX showed that the biological process associated with this cluster includes \"cell gliding\" (GO:0071976), \"vacuolar transport\" (GO:0007034), \"Cell motility\" GO:0048870, \"regulation of actin filament length\" (GO:0030832), \"entry into host cell\" (GO:0030260), \"protein palmitoylation\" (GO:0018345). There was no GO term that was significantly (p > 0.05) associated with cluster 5. The significant processes however may be of some importance in this life-cycle stage. Previous investigations have demonstrated that invasion is dependent on the gliding and motility of the parasite which is dependent on the actin and myosin present in the parasite's pellicle (Boysen & Matuschewski, 2013; Dobrowolski et al., 1997; Dobrowolski & Sibley, 1996; Heintzelman & Schwartzman, 1997). Invasion of the parasite also involves vacuolar transportation by rhoptries. Rhoptries are club-shaped structures containing a long duct through which the luminal contents are extruded at the time of host cell invasion. Thought to be similar in function to multi-vesicular bodies (MVBs), they may function in sorting specific proteins and lipids. Unlike MVBs however, their membranes are characterized by an unusually high cholesterol – phospholipid ratio (Sonda & Hehl, 2006; Yang et al., 2004). The requirement of these three processes for invasion is a reasonable explanation for their coordinated regulation. The sustained expression till the end of the gametocyte stage may also suggest that key genes involved in the processes are essential to the survival of the parasite at the gametocyte – gamete stage or at least, are important in the events leading to this transition which may be potential candidates for transmission drug targeting.\n\nOnly the mature gametocytes can undergo further development in the mosquito mid-gut. Gametocytes escape the red blood cells taken up in the blood meal by the mosquito to form gametes - the male gametocyte differentiating into eight sperm-like gametes involving three rounds of DNA replication and nuclear division which happens in the first 10 minutes. The female differentiates into a single spherical gamete. This is triggered by environmental factors, which include the drop in temperature, the rise in pH and the presence of activating factors in the new vector host. Between 10 minutes and 1 hour, fertilization occurs (Landau & Gautret, 1998). Meiosis does not immediately occur after nuclear division resulting in an ookinete with a higher amount of DNA. This ookinete develops into a spherical, more motile ookinete with an apical complex for traversing the mid-gut epithelium after 24 hours (Landau & Gautret, 1998). In MOS, 2 clusters of biological interest were identified. Clusters 4 (153 genes) showed induction of genes only after 16 hours. GO terms associated with this cluster include \"peptidyl-arginine N-methylation\" (GO:0035246), \"regulation of cell proliferation\" (GO:0042127), \"protein processing involved in protein targeting to mitochondrion\" (GO:0006627), \"protein farnesylation\" (GO:0018343) and \"U6 snRNA 3'-end processing\" (GO:0034477). Peptidyl-arginine N-methylation has been shown to regulate cellular process including RNA processing, transcriptional regulation, signal transduction, DNA repair and also plays a role in targeting proteins to the plastid. There are interesting results from studies employing peptidyl-arginine N-methylation inhibitors that call for further investigation for a drug strategy in malaria (Bedford & Richard, 2005; Dillon et al., 2012). Protein farnesylation involves lipid post-translational modification that occurs in eukaryotic cells. In higher eukaryotes, farnesylation of some proteins (notably GTPase Ras) play a role in cell signal transduction, vesicle trafficking, and cell cycle progression. Increased levels of these proteins can potentially lead to cancer for which inhibitors (PFTIs) have been designed. Investigators are currently researching the use of PFTIs for against eukaryotic pathogens including Plasmodium species (Eastman et al., 2006; Tamanoi et al., 2001; Wiesner et al., 2004). 3'-end processing of snRNA and ncRNA are important in RNA biogenesis. Genes in cluster 9 (258 genes) were induced between the 0 – 16 hour with a rapid decline in trend after this. GO terms associated with these clusters include \"peptidyl-diphthamide biosynthetic process from peptidyl-histidine\" (GO:0017183) and \"vesicle docking involved in exocytosis\" (GO:0006904). Diphtamide is a unique post-translationally modified histidine residue of the elongation factor 2 conserved in eukarotic cells. Although the biological function of diphtamide is still not clearly understood, diphtamide has a well-studied pathological application. It functions as recognition site by some toxins (Su et al., 2013). How diphtamide makes cells susceptible to exploitation by toxins should be investigated. The triggering of calcium-mediated signaling pathways within the sporozoite is thought to be necessary to induce exocytosis of molecules required for sporozoite invasion (Carey et al., 2014). To date, no specific Plasmodium calcium dependent serine/threonine protein kinase (CDPK) inhibitors have been discovered even though seemingly abundant in the genome.\n\nCis-regulatory elements are generally known to be located in regions upstream of the gene start codons. In this study, we searched for candidate motifs up to 800 bases upstream of each gene set for matching sequences using three different algorithms. From previous motifs that have been experimentally determined (Painter et al., 2011) a reliable candidate motif in the AT-rich and repetitive Plasmodium genome should have a reasonable degree of variability containing at least two or more bases other than A or T with a low occurrence in other locations in the Plasmodium genome (Young et al., 2008). Our study shows five candidate motifs for cis-regulatory elements obtained after analysis of clusters of each of the developmental stages of the parasite. The attributes of each motif is illustrated by the position weight matrices (PWM) logos, one of which is corroborated by an independent research carried out by a separate set of scientists employing a different method in Figure 2. As previously discussed, clustering of genes with similar expression patterns revealed 14 biological processes and pathways which possibly played integral roles in distinct stages of its developmental life-cycle of the Plasmodium species in response to transduced stimuli that lead to regulation of gene expression in the nucleus as a result of changes in its environments. The paucity of identified regulatory elements in Plasmodium is largely attributed to the nature of its AT-rich and highly repetitive genome. This will suggest that genes involved in these otherwise unrelated processes possibly share common regulatory sequences. This provides a possible explanation to how Plasmodium can coordinate complex responses to environmental changes with limited regulatory routes.\n\nSignificant values > 0 are relevant. References of logos identified in other research using other methods are highlighted in column 5.\n\n\nConclusion\n\nIn this study analyzed Plasmodium RNA-seq data covering three life developmental stages to identify biological processes as possible candidates for drug targeting and their respective points of coordinated transcription control. To achieve this goal, we used an unsupervised machine learning approach, grouping genes which showed similar expression patterns across time-point for each respective developmental stage of the parasite. We showed that each development stage activated biological process custom to the anticipated environment unique to each development stage in which we identified 14 biological processes that may be integral at different stages of the parasites development, 11 of which have not been investigated in drug-related research. These include sustained upregulated biological processes linked to invasion e.g. peptidyl-diphathamine biosysthesis and protein processes involved in mitochondria targeting which could be targeted when designing drugs that prevent transmission. In agreement with other studies, the IDC expressed the fewest genes, a possible survival adaptation of the parasite, which makes targeting that stage difficult. In addition to these, some uncharacterized genes were also identified that may yet yield new drug targets following further investigations. We also showed in our study that genes involved in more than one biological process may share common regulatory elements which may explain the paucity of transcription factors unique to the species. We elucidated five such consensus sequences, four of which are novel sequences that may be potential cis-regulatory elements involved in coordinated control pending further validation studies.\n\n\nData availability\n\nData files are openly available at https://github.com/efejiroe/pb-tc-cluster-experiment-suppdata.\n\nF1000Research: Dataset 1. Data of cluster analysis of Plasmodium RNA-seq time-course data identifying stage-specific co-regulated biological processes and regulatory elements, 10.5256/f1000research.9093.d131414 (Ashano et al., 2016).",
"appendix": "Author contributions\n\n\n\nAshano Efejiro and Isewon Itunuoluwa conceived the study and retrieved the data. Ashano Efejiro performed the in silico analyses and interpretation of the data. Ashano Efejiro and Isewon Itunuoluwa prepared the initial draft of the manuscript. Adebiyi Ezekiel and Oyelade Jelili further improved the manuscript. All authors read and approved the final contents of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAllary M, Lu JZ, Zhu L, et al.: Scavenging of the cofactor lipoate is essential for the survival of the malaria parasite Plasmodium falciparum. Mol Microbiol. 2007; 63(5): 1331–1344. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAshano E, Isewon I, Oyelade J, et al.: Dataset 1 in: Cluster analysis of Plasmodium RNA-seq time-course data identifies stage-specific co-regulated biological processes and regulatory elements. F1000Research. 2016. Data Source\n\nBaum J, Gilberger TW, Frischknecht F, et al.: Host-cell invasion by malaria parasites: insights from Plasmodium and Toxoplasma. Trends Parasitol. 2008; 24(12): 557–563. PubMed Abstract | Publisher Full Text\n\nBedford MT, Richard S: Arginine methylation an emerging regulator of protein function. Mol Cell. 2005; 18(3): 263–272. PubMed Abstract | Publisher Full Text\n\nBenjamini Y, Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B. 1995; (57): 289–300. Reference Source\n\nBoysen KE, Matuschewski K: Inhibitor of cysteine proteases is critical for motility and infectivity of Plasmodium sporozoites. MBio. 2013; 4(6): e00874–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBozdech Z, Llinás M, Pulliam BL, et al.: The transcriptome of the intraerythrocytic developmental cycle of Plasmodium falciparum. PLoS Biol. 2003; 1(1): E5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCallebaut I, Prat K, Meurice E, et al.: Prediction of the general transcription factors associated with RNA polymerase II in Plasmodium falciparum: conserved features and differences relative to other eukaryotes. BMC Genomics. 2005; 6(1): 100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarey AF, Singer M, Bargieri D, et al.: Calcium dynamics of Plasmodium berghei sporozoite motility. Cell Microbiol. 2014; 16(5): 768–783. PubMed Abstract | Publisher Full Text\n\nCarlson JM, Chakravarty A, DeZiel CE, et al.: SCOPE: a web server for practical de novo motif discovery. Nucleic Acids Res. 2007; 35(Web Server issue): W259–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDillon MB, Bachovchin DA, Brown SJ, et al.: Novel inhibitors for PRMT1 discovered by high-throughput screening using activity-based fluorescence polarization. ACS Chem Biol. 2012; 7(7): 1198–1204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDobrowolski JM, Carruthers VB, Sibley LD: Participation of myosin in gliding motility and host cell invasion by Toxoplasma gondii. Mol Microbiol. 1997; 26(1): 163–173. PubMed Abstract | Publisher Full Text\n\nDobrowolski JM, Sibley LD: Toxoplasma invasion of mammalian cells is powered by the actin cytoskeleton of the parasite. Cell. 1996; 84(6): 933–939. PubMed Abstract | Publisher Full Text\n\nEastman RT, Buckner FS, Yokoyama K, et al.: Thematic review series: lipid posttranslational modifications. Fighting parasitic disease by blocking protein farnesylation. J Lipid Res. 2006; 47(2): 233–240. PubMed Abstract | Publisher Full Text\n\nEisen MB, Spellman PT, Brown PO, et al.: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A. 1998; 95(25): 14863–14868. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEjigiri I, Ragheb DR, Pino P, et al.: Shedding of TRAP by a rhomboid protease from the malaria sporozoite surface is essential for gliding motility and sporozoite infectivity. PLoS Pathog. 2012; 8(7): e1002725. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreenwood BM, Fidock DA, Kyle DE, et al.: Malaria: progress, perils, and prospects for eradication. J Clin Invest. 2008; 118(4): 1266–1276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGünther S, Storm J, Müller S: Plasmodium falciparum: organelle-specific acquisition of lipoic acid. Int J Biochem Cell Biol. 2009; 41(4): 748–752. PubMed Abstract | Publisher Full Text\n\nHarris EY, Ponts N, Le Roch KG, et al.: Chromatin-driven de novo discovery of DNA binding motifs in the human malaria parasite. BMC Genomics. 2011; 12(1): 601. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeintzelman MB, Schwartzman JD: A novel class of unconventional myosins from Toxoplasma gondii. J Mol Biol. 1997; 271(1): 139–146. PubMed Abstract | Publisher Full Text\n\nHerrero J, Valencia A, Dopazo J: A hierarchical unsupervised growing neural network for clustering gene expression patterns. Bioinformatics. 2001; 17(2): 126–136. PubMed Abstract | Publisher Full Text\n\nJurgelenaite R, Dijkstra TM, Kocken CH, et al.: Gene regulation in the intraerythrocytic cycle of Plasmodium falciparum. Bioinformatics. 2009; 25(12): 1484–91. PubMed Abstract | Publisher Full Text\n\nLandau I, Gautret P: Animal models: rodents. Malaria: Parasite Biology Pathogenesis and Protection. 1998; 401–417.\n\nLukashin AV, Fuchs R: Analysis of temporal gene expression profiles: clustering by simulated annealing and determining the optimal number of clusters. Bioinformatics. 2001; 17(5): 405–414. PubMed Abstract | Publisher Full Text\n\nMcGarvey MJ, Sheybani E, Loche MP, et al.: Identification and expression in Escherichia coli of merozoite stage-specific genes of the human malarial parasite Plasmodium falciparum. Proc Natl Acad Sci U S A. 1984; 81(12): 3690–3694. PubMed Abstract | Free Full Text\n\nNueda MJ, Tarazona S, Conesa A: Next maSigPro: updating maSigPro bioconductor package for RNA-seq time series. Bioinformatics. 2014; 30(18): 2598–602, btu333. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOtto TD, Böhme U, Jackson AP, et al.: A comprehensive evaluation of rodent malaria parasite genomes and gene expression. BMC Biol. 2014; 12(1): 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPainter HJ, Campbell TL, Llinás M: The Apicomplexan AP2 family: integral factors regulating Plasmodium development. Mol Biochem Parasitol. 2011; 176(1): 1–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRao MR: Foreword: International Centers of Excellence for Malaria Research. Am J Trop Med Hyg. 2015; 93(3 Suppl): 1–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSarker M, Talcott C, Galande AK: In silico systems biology approaches for the identification of antimicrobial targets. Methods Mol Biol. 2013; 993: 13–30. PubMed Abstract | Publisher Full Text\n\nSims JS, Militello KT, Sims PA, et al.: Patterns of gene-specific and total transcriptional activity during the Plasmodium falciparum intraerythrocytic developmental cycle. Eukaryotic Cell. 2009; 8(3): 327–338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSonda S, Hehl AB: Lipid biology of Apicomplexa: perspectives for new drug targets, particularly for Toxoplasma gondii. Trends Parasito. 2006; 22(1): 41–47. PubMed Abstract | Publisher Full Text\n\nStorm J, Müller S: Lipoic acid metabolism of Plasmodium--a suitable drug target. Curr Pharm Des. 2012; 18(24): 3480–3489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSu X, Lin Z, Lin H: The biosynthesis and biological function of diphthamide. Crit Rev Biochem Mol Biol. 2013; 48(6): 515–521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSubramaniam PS, Torres BA, Johnson HM: So many ligands, so few transcription factors: a new paradigm for signaling through the STAT transcription factors. Cytokine. 2001; 15(4): 175–187. PubMed Abstract | Publisher Full Text\n\nSultan AA, Thathy V, Frevert U, et al.: TRAP is necessary for gliding motility and infectivity of Plasmodium sporozoites. Cell. 1997; 90(3): 511–522. PubMed Abstract | Publisher Full Text\n\nTamanoi F, Gau CL, Jiang C, et al.: Protein farnesylation in mammalian cells: effects of farnesyltransferase inhibitors on cancer cells. Cell Mol Life Sci. 2001; 58(11): 1636–1649. PubMed Abstract | Publisher Full Text\n\nTarca AL, Romero R, Draghici S: Analysis of microarray experiments of gene expression profiling. Am J Obstet Gynecol. 2006; 195(2): 373–388. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Z, Gerstein M, Snyder M: RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009; 10(1): 57–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiesner J, Kettler K, Sakowski J, et al.: Farnesyltransferase inhibitors inhibit the growth of malaria parasites in vitro and in vivo. Angew Chem Int Ed Engl. 2004; 43(2): 251–254. PubMed Abstract | Publisher Full Text\n\nYang M, Coppens I, Wormsley S, et al.: The Plasmodium falciparum Vps4 homolog mediates multivesicular body formation. J Cell Sci. 2004; 117(Pt 17): 3831–3838. PubMed Abstract | Publisher Full Text\n\nYoung JA, Johnson JR, Benner C, et al.: In silico discovery of transcription regulatory elements in Plasmodium falciparum. BMC Genomics. 2008; 9: 70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Y, Young JA, Santrosyan A, et al.: In silico gene function prediction using ontology-based pattern identification. Bioinformatics. 2005; 21(7): 1237–1245. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15551",
"date": "14 Sep 2016",
"name": "Manuel Corpas",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article \"Cluster analysis of Plasmodium RNA-seq data identifies stage-specific co-regulated biological processes and regulatory elements\" uses an RNA-seq dataset published in Otto et al. 1 for a number of Plasmodium species. Authors use the assumption that co-transcribed genes are likely to be controlled by a common regulatory element. Subsequently they provide a clustering of expression profiles using that dataset.\nI would propose the following enhancements to the article:\nIn order to ensure the reproducibility of the experiment, I would request if more information could be provided about the raw dataset from which the study has derived. Things like number of samples, what conditions, how big the data is, etc. This information should be available in the source paper.\n\nWhat parameters did you use to run maSigPro? What parameters did you use to cluster gene expression patterns?\n\nI am concerned about the use of Herrero et al.'s tool, which was published in 2001. This tool was developed for array data, a long time before NGS data like transcriptomics became available.\n\nIn results: 155 genes out of 1448 is not ~0.1%. It's ~10%.\n\nIn table 2, I am not sure what the background percentages are. I would appreciate if legends made figures self explanatory.\n\nWhen presenting results, for each section there is a lot of background information. For example:\nSection 1: \"Characterized genes may provide alternative sources for drug targets of the IDC\". I find this header too obvious and uninformative. The text at the beginning of this section up to the sentence \"For the ICD, to clusters [...]\" is all material that would be better placed in a discussion section. Clearly this text is not part of what I would call \"results\".\nI find the other sections within Results also have a lot of preliminary background information that in my opinion would be more suitable for a discussion section than results. Another example is the section \"Sustains up-regulated [...]\". I still think this background information is necessary though to understand the context of results.\n\nI would find it helpful if Figure 3 was referenced from the text in the relevant section. Currently, as it appears in the paper, I had no idea why of how it related to the paper until I read the section \"Sustained up-regulated biological processes linked to invasion may present key drug targets to prevent transmission\".\n\nIn conclusions I don’t understand this sentence: “We showed that each development stage activated biological process custom to the anticipated environment unique to each development stage in which we identified 14 biological processes that may be integral at different stages of the parasites development, 11 of which have not been investigated in drug-related research.\"",
"responses": []
},
{
"id": "17942",
"date": "24 Nov 2016",
"name": "Hilary Ann Coller",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscripts describes a carefully executed analysis of gene expression patterns over the Plasmodium life cycle. The authors have identified genes and GO categories that are upregulated at different stages of the Plasmodium life cycle, the intraerythrocytic developmental cycle, the gamete-fertilization stage and the sexual development stage. The genes that are upregulated suggest some pathways that may be targeted as new drugs to combat malaria. The authors also identified five motifs that are enriched in the promoters of the genes regulated at the different stages. Identifying enriched motifs is especially challenging in these species because of their AT-rich genomes. The title and abstract are accurate reflections of the work. The experimental design is careful and well-described. The conclusions are appropriate. The work relies on data that are already in the public domain.",
"responses": []
},
{
"id": "17785",
"date": "09 Dec 2016",
"name": "Thomas D. Otto",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary\n\nAshano et al apply a clustering method for a reanalysis of a published dataset. The dataset is split into three groups and in each they identify genes that exhibit significant upregulation of gene expression and then cluster genes by RNA-seq gene expression profile. They identify clusters that exhibit stage specific upregulation and show that certain clusters have enriched GO terms associated with processes specific to the life cycle stage. Finally, they search for sequence motifs that are enriched in the upstream region of genes in a specific cluster, suggesting that these motifs may be regulatory elements for the clusters.\n\nReview Comments\n\nThe main criticism is the difficulty to understand the clustering, which is central to the paper. We believe that the clustering may contain fundamental flaws that are difficult to pinpoint due to the unclarity of the methods.\nIt seems that the six different time points are split into three groups IDX, SEX and MOS. This is unintuitive and discards genes expressed at different stages.\n\nHow robust is the clustering using maSigPro to different parameters?\n\nHow was 9 clusters chosen?\n\nDoes a different clustering algorithm yield similar results?\n\nThe explanation of how genes are determined as being significantly upregulated and how the clustering is performed is unclear at the moment. Some more explanation would be useful. Particularly, it is not clear if only significantly upregulated genes are clustered or not, as later in the manuscript the authors refer to clusters with significantly upregulated genes suggesting that non-significant genes are also clustered.\n\nBetter visualization of the clusters would be nice. For example, how many clusters are there and how many genes are in each of the clusters? In figure 1, it would be good to plot the expression profiles of all the genes in each of the clusters in order to get an idea of the variance in gene expression in each of the clusters.\n\nWhen describing the proportion of genes upregulated in each of the life-cycle stages, it would be useful to compare these results to other published datasets to look at the overlap.\nTherefore the clustering results are unexpected. Figure 1 should show all the time points. We are not sure what the 0h time point refers to. Further just 1448 genes seem to be differentially expressed in the IDX, but this is in contrast to around 3300 genes in the IDX expressed in several P. falciparum studies1-3.\nAs the title of the manuscript suggests an analysis of Plasmodium, why choose a dataset that only has 3 idx timepoints, compared to several other RNA-Seq datasets with at least 7 timepoints, or over 48 timepoints in the above mentioned microarray experiments. There are a lot of published data, see plasmoDB. So it is unclear why this choice was made. In the absence of a rational reason, it would be good to use an alternative dataset to get better resolution.\n\nLast, we do not understand how 4383 genes out of 5003 can be differentially expressed in the MOS stage, Table 1 of the manuscript. In the figure 2 of the paper presenting the used date4, the amount of genes expressed in the ookinete stage is similar to that of the other stages. Could you please explain the differences. Better data visualization and more explicit numbers quoted in a more thorough walkthrough of the methods would help clear up these questions.\n\nIf the authors choose not to include further datasets, the title or at least the abstract should specify that the analysis was done on P. berghei. The manuscript currently generalizes the findings in this species to be generally valid for all other Plasmodium species.\n\nThe introduction is very general, and could give a better overview of similar work performed in malaria recently.\n\nIn the GO enrichment, genes that are not expressed in the dataset should be excluded from the gene background (which is not possible in plasmoDB). It would be better to use topGO. Also, the GO terms are from geneDB and not from the annotation center.\n\nWhy mention SEX cluster 5 in the GO term section if there are no significant GO terms?\n\nUpstream regions of genes for the motif identification were obtained from P. falciparun orthologs. Why not use P. berghei upstream regions, seeing as the RNA-seq data is from P. berghei?\n\nIn the results it is stated that motifs were searched in the 800bp upstream region of each gene, but in the methods it says 1000bp. Which is correct?\n\nAdditional information on the identified motifs would be nice. For example, distribution of the motifs in the upstream region, presence of the motifs in upstream regions of other genes, etc. Additional comparison of the motifs to motifs identified in other studies would also be good, for example known AP2 binding motifs, with a lot of work having been performed by Ilinas 5.\n\nIt is highly speculative to state that a lower number of genes are upregulated in the IDC stage due to immune pressure. The mosquito also possesses an immune system that would likely exhibit similar pressure.\n\nIn table 2, the background percent column would benefit from better explanation and explicit numbers: ie. 5 genes out of 10 instead of just 50%. P-values might also be useful to include here.\n\nIn many cases references are placed at the end of paragraphs instead of after the particular statement that the references support, for example in the lipoate paragraph.\n\nLast, we would encourage the authors to discuss their finding with other similar papers in the field, especially the found motifs and the upregulated gene clusters.\n\nThere are multiple spelling mistakes and strangely formulated sentences throughout the manuscript that hinder understanding.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1932
|
https://f1000research.com/articles/5-1923/v1
|
05 Aug 16
|
{
"type": "Research Article",
"title": "Neurosteroids are reduced in diabetic neuropathy and may be associated with the development of neuropathic pain",
"authors": [
"Stephen R. Humble"
],
"abstract": "Introduction: Peripheral and central sensitisation are implicated in the development of neuropathic pain. Hypersensitivity of pain pathway neurons has been described in animal models of diabetic neuropathy, which is postulated to be related to an imbalance between inhibitory and excitatory signals within the spinal cord. GABAergic neurons within the pain pathway are vital for the transmission of painful stimuli to higher centres. A developmental change in the rate of exponential decay of GABAergic synaptic events has been observed in other types of neurons and this may be associated with fluctuations in endogenous neurosteroid tone.\n\nMethods: The whole-cell patch-clamp technique was used on slices of neural tissue. Electrophysiological recordings were obtained from wild type mice between the ages of 6 and 80 days in the spinal cord, the nucleus reticularis of the thalamus and the cerebral cortex. Recordings were also obtained from mice with diabetic neuropathy (ob/ob and db/db) between the ages of 60 and 80 days. Behavioural experiments were performed to examine mechanical and thermal nociception.\n\nResults: Electrophysiological recordings from cortical pain pathway neurons from mature type-2 diabetic mice revealed that the endogenous neurosteroid tone is reduced compared to control. However, selected neurosteroid compounds had a more pronounced effect on the GABAA receptors of these diabetic mice. ob/ob mice exhibit mechanical hyperalgesia and allodynia, which was reduced by neurosteroids applied exogenously.\n\nConclusions: The reduced endogenous neurosteroid tone in ob/ob mice may be linked to their hypersensitivity. Neurosteroids may exert analgesic effects in pathological pain states by attempting to restore the physiological GABAergic inhibitory tone.",
"keywords": [
"Neuropathic pain",
"Neurosteroids",
"diabetes",
"neuropathy",
"GABAAR",
"GABAA receptor",
"ob ob",
"db db"
],
"content": "\n\n\n\nThere is a worldwide obesity diabetes epidemic, causing a huge amount of morbidity including neuropathy and neuropathic pain.\n\nThe mechanisms responsible for the development of neuropathy and neuropathic pain are not well understood.\n\nPeripheral and central sensitisation are implicated in the development of neuropathic pain with neuroplastic changes occurring at multiple levels of the pain pathway.\n\nHypersensitivity of pain pathway neurons has been described in animal models of diabetic neuropathy, which is postulated to be related to an imbalance between inhibitory and excitatory signals within the spinal cord. GABAergic neurons within the pain pathway are vital for the transmission of painful stimuli to higher centres involved in the perception of pain.\n\nGABAA receptors are an important target for many drugs, and specific endogenous neurosteroids act as potent allosteric modulators of these receptors.\n\nA developmental change in the rate of exponential decay of GABAergic synaptic events has been observed in other types of neurons and this may be associated with fluctuations in endogenous neurosteroid tone.\n\nThis paper explores potential underlying mechanisms and identifies potential therapeutic targets in order to promote translational work in this field.\n\nThis is the first electrophysiological GABAA receptor characterisation of two models of type-2 diabetes mellitus and reports the discovery of a reduced endogenous GABA-ergic neurosteroid tone that may mediate painful neuropathic hypersensitivity.\n\nThe paper then reports the anti-nociceptive impact of neurosteroids on live mice, thus complementing the electrophysiological data.\n\nThese discoveries may ultimately lead to a new rational avenue of research aimed at understanding and treating painful diabetic neuropathy and the neuropathic mechanisms may be analogous in other neuropathic conditions such as chemotherapy-induced neuropathic pain.\n\n\nIntroduction\n\nA loss of physiological inhibitory tone is associated with hypersensitivity to painful stimuli such as allodynia and hyperalgesia (Chen & Pan, 2002; Zeilhofer, 2008). The GABAA receptor (GABAAR) is the major inhibitory receptor in the mammalian nervous system and mediates inhibitory tone throughout the pain pathway (D’Hulst et al., 2009; Johnston, 2005). Reduced GABAergic inhibition within the spinal cord may be implicated in the development of hypersensitivity to nociceptive stimuli (Munro et al., 2009; von Hehn et al., 2012; Zeilhofer, 2008). Therefore, pharmacological agents that enhance GABAAR function could be useful to counteract lost inhibitory tone (Knabl et al., 2008; Munro et al., 2009). Neurosteroids such as allopregnanolone are potent allosteric modulators of this receptor (Callachan et al., 1987; Hosie et al., 2006; Figure 1). Indeed, an upregulation in the production of endogenous neurosteroids within the spinal cord in response to peripheral inflammation has been shown to have an analgesic effect. The analgesic effect could be suppressed by the administration of finasteride to inhibit the enzyme 5α-R, which converts progesterone to its more active metabolites (Poisbeau et al., 2005; Schlichter et al., 2006).\n\nCholesterol is taken through the mitochondrial membrane by the translocator protein where it is converted to pregnenolone by the cytochrome P450 side chain cleavage enzyme. Pregnenolone undergoes stepwise enzymatic conversion to other neurosteroid compounds and ultimately allopregnanolone which modulates GABAAR function. Neurosteroids may act via paracrine or autocrine mechanisms. Postsynaptic GABAARs are activated by GABA that has been released from vesicles in the presynaptic nerve terminal. GABA induces a conformational change in the GABAAR, which ‘opens’ its central pore, allowing the passage of chloride ions. The negatively charged chloride ions typically induce hyperpolarisation of the neuronal membrane, which is associated with neuronal inhibition. Neurosteroids such as allopregnanolone enhance GABAAR function and therefore facilitate neural inhibition.\n\nThere is a burgeoning obesity and type-2 diabetes epidemic worldwide and the incidence is likely to become even greater (Danaei et al., 2011). Neurosteroids such as progesterone have been studied for their potentially protective effects for a number of different neuropathologies including stroke, brain and spinal injuries (Mensah-Nyagan et al., 2009; Stein, 2008). It is possible that fluctuations in endogenous neurosteroid levels may have a pathophysiological role in the development of diabetic neuropathic pain. In mice, diabetic neuropathy develops over weeks, which allows disease progression and the efficacy of interventions to be studied within a relatively a short timeframe (Cefalu, 2006; Kaplan & Wagner, 2006). The type-2 diabetic ob/ob mouse, which has an autosomal recessive nonsense mutation on chromosome 6 causing leptin deficiency develops morbid obesity due to its rapacious appetite and exhibits a predictable and spontaneous neuropathic phenotype (Drel et al., 2006; Latham et al., 2009; Lindstrom, 2007; Vareniuk et al., 2007). In contrast, the db/db mouse has an autosomal recessive mutation of the leptin receptor gene, which prevents leptin from activating its receptor (Chen et al., 1996; Chung et al., 1996).\n\nPain has sensory, emotional and cognitive components and therefore multiple areas within the cerebral cortex are involved in generating the experience of pain (Flor & Bushnell, 2005; Treede et al., 1999). Areas such as the insular, somatosensory, anterior cingulate and prefrontal cortices and the thalamus may be considered as a ‘pain matrix.’ The cerebral cortex is organised into parallel mini-columns of synaptically linked neurons that span part, or all of the six horizontal cortical layers. Mini-columns may be clustered together to constitute functional modules (Lubke & Feldmeyer, 2007; Mountcastle, 1997). GABAergic inhibitory interneurons are present in all layers of the cortex but are most abundant in layers 2/3 and lower in layer 4 (Meyer et al., 2011). By improving the understanding at the molecular level it may be possible to identify novel therapeutic targets for painful diabetic neuropathy.\n\n\nMethods\n\nAll procedures were carried out in compliance with the University of Dundee code of practice following consideration by the University Research Ethics Committee (UREC) and in accordance with Schedule 1 of the Animals (Scientific Procedures) Act 1986 (UK). Home Office Project Licence numbers: PPL 60/4144 and PPL 60/4005. Wellcome Trust Grant number 090667. Wild Type (WT) mice aged less than 2 months were obtained from an in-house colony. C57/Bl6J WT mice aged 2 months were purchased from Charles River, UK while OlaHSD ob/ob mice, db/db mice and their respective strain-matched WT littermates, all aged 2 months were purchased from Harlan UK and housed in the same in-house colony. All animals were male and were kept under an alternating 12hr light/dark cycle and had ad libitum access to food and water.\n\nPrior to dissection mice were killed instantly by cervical dislocation (Schedule 1), the tissue was then submerged immediately into a bath of ice-cold oxygenated (95% O2/5% CO2) artificial cerebrospinal fluid (aCSF). The aCSF comprised: 225mM sucrose, 10mM glucose, 10mM MgSO4, 26mM NaHCO3, 1.25mM NaH2PO4, 2.5mM KCl, 0.5mM CaCl2, 1mM ascorbic acid and 3mM pyruvic acid (total osmolarity ~330mosm/l). The spinal cord was extracted by anterior laminectomy as per Keller et al. (2001). The cord was immediately set within agar gel, which was glued to the Leica VT1000S vibratome (Heidelberg, Germany) for slicing. Horizontal, thoracolumbar slices, (300–350μm) were cut and transferred onto a nylon mesh platform within a storage chamber containing oxygenated (95% O2/5% CO2) artificial extracellular solution (aECS) comprising: 126mM NaCl, 10mM glucose, 2mM MgCl2, 26mM NaHCO3, 1.25mM NaH2PO4, 2.95mM KCl, 2mM CaCl2, 1mM ascorbic acid, 3mM pyruvic acid and 2mM kynurenic acid (Total measured osmolarity ~310mOsm/l). Slices were stored at room temperature for at least one hour before being used to obtain electrophysiological recordings.\n\nBrain tissue was also obtained following the cervical dislocation method (Schedule 1) and submerged in aCSF as above. For all nucleus reticularis (nRT) preparations the aCSF solution was the same as that described above for spinal cord but the sucrose concentration was increased to 234mM giving a total osmolarity of ~340mOsm/l as this improved the condition of the slices. For cortical preparations in mice below 2 months of age the aCSF was the same as for the spinal cord. Neuronal viability deteriorates with increasing age and for mice above the age of 2 months, including all ob/ob and db/db, a different solution was required to optimise the condition of the slices as previously described by Maguire et al. (2014). This consisted of 140mM potassium gluconate, 10mM HEPES, 15mM sodium gluconate, 0.2mM EGTA, 4mM NaCl, 1mM ascorbic acid and 3mM pyruvic acid. Sodium hydroxide solution was then added to bring the pH up to 7.2. The brain was removed carefully from the skull and slices were obtained using a Vibratome series 1000 PLUS Sectioning System (Intracell, Royston, Hertfordshire, UK). Cortical slices were cut in the coronal plane and nRT slices in the horizontal plane. Slice thickness was greatest for the youngest mice and least for the oldest mice (250–350μM). Slices were transferred immediately to a storage chamber as previously described.\n\nSubsequently, the slices were transferred to a recording chamber under an Olympus BX51WI fixed-stage upright microscope. An infrared differential interference contrast disc and a water immersion objective (× 40) were used for visualisation of neurons within the dorsal horn of the spinal cord, the nucleus reticularis of the thalamus and the somatosensory area of the cortex respectively. The chamber was perfused continuously with oxygenated artificial extracellular solution (aECS). Slices were held in position using a small grid. Whole-cell patch-clamp recordings were obtained using an Axopatch 200B amplifier and all recordings were made at a holding potential of -60mV and at a temperature of 35°C. The extracellular solution contained kynurenic acid (2mM), strychnine (0.5μM) and tetrodotoxin (0.5μM) to antagonise ionotropic glutamate, glycine receptors and voltage-gated sodium channels respectively. Patch electrodes were prepared from thick walled borosilicate glass capillaries (Garner Glass Co., Claremont, CA, USA). Such electrodes had an open-tip resistance of ~4MΩ when filled with intracellular solution containing: 135mM CsCl, 10mM HEPES, 10mM EGTA, 1mM CaCl2, 1mM MgCl2, 2mM Mg-ATP, 5mM QX-314, pH 7.2, titrated with CsOH. The measured osmolarity was 310mOsmol/l. Typically, the mean whole-cell capacitance was 5–15pF. Series resistance was compensated for by up to 80% and recordings were considered invalid where the series resistance changed by more than 20% or if it exceeded 15MΩ. A 2 kHz frequency filter was used for all recordings and analysis of each cell was performed offline.\n\nThe Strathclyde Electrophysiology Software, WinEDR and WinWCP, (Dr J Dempster, University of Strathclyde, Glasgow, UK) was used for the analysis of recordings. Only recordings that met specific quality criteria were included in the analysis. Using an algorithmic detection protocol, miniature inhibitory postsynaptic currents (mIPSCs) were detected with an amplitude threshold of at least -5pA and duration > 2ms. Each individual mIPSC was then inspected visually to ensure validity and exclude artifactual events. mIPSCs with a rise time of more than 1ms were excluded to prevent the inclusion of events originating from a distal source. At least 50 events were sought for each recording, and the average peak amplitude, rise time (10–90%), charge transfer (area under the curve) and T70 (time required to decay by 70%), were analysed. The mIPSCs were digitally averaged by alignment at the midpoint of the rising phase. The mIPSC decay was fitted by monexponential (y(t) = Ae-t/τ) and biexponential (y(t) = Afaste(-t/τfast)+ A slowe(-t/τslow)) functions to determine which one was more appropriate (A is amplitude, t is time and τ is the decay time constant). The standard deviations of the residuals of the monexponential and biexponential functions were measured and an F test applied. For the vast majority of mIPSCs analysed, the decay was best described by a biexponential function. Consequently, a mean weighted decay constant (τw) was calculated to determine the relative contribution of each decay component. τw is a mathematical constant generated by considering the fast initial component of biexponential decay τ1 and the later slower component τ2. The value of τw is determined for the mean mIPSC for each cell by determining the relative proportion that τ1 and τ2 contribute to the biexponential decay. The following equation was used: τw = τfastP1 + τslowP2, where P1 and P2 represent the proportions of the synaptic current decay curve described by each component.\n\nSalts used in the preparation of aECS and aCSF solutions were purchased from VWR (West Chester, Pennsylvania, USA). Strychnine (Sigma Chemicals, St. Louis, MO, USA), tetrodotoxin (Tocris, Bristol, UK) and bicuculline (Axxora, Nottingham, UK) and THIP (generous gift from B Ebert) were prepared as concentrated stock solutions in double-filtered water to be added to the aECS. Other compounds such as progesterone, ganaxolone, allopregnanolone, dihydroxy-progesterone, provera, indometacin, finasteride were purchased from Tocris or Sigma and prepared as concentrated stock solutions in dimethyl sulfoxide (DMSO). The cyclodextrins (CDs; Sigma) were dissolved directly into the aECS, or the intracellular solution. Brain and spinal cord slices were placed under the microscope into a transparent plastic chamber filled with aECS. The aECS was perfused through the chamber using a rate-adjustable gravity-based system consisting of hard plastic tubing that connected an oxygenated reservoir to the chamber. Simultaneously, a peristaltic pump system (Minipuls 3, Gilson, UK) drained the aECS from the opposite side of the chamber and recycled it back to the reservoir. The fluid-filled circuit also ran through a custom made heating system (G23, UCL, London, UK) controlled by a temperature monitoring system (School of Pharmacology, London, UK) that maintained the near physiological temperature of 35°C throughout.\n\nDrug administration. The drug, or vehicle, was administered by intra-peritoneal injection using a fine 1ml syringe and experiments were made pre- and post- injection in a randomised and blinded manner. Neurosteroids are lipophilic and have very limited solubility in aqueous solution, but can be solubilised in 0.9% saline using 2-hydroxy propyl β-cyclodextrin (β-CD) to facilitate administration (Besheer et al., 2010; Carter et al., 1997; Reddy & Rogawski, 2010). Neurosteroids were therefore administered in 40% β-CD solution.\n\nRotarod test. The rotarod test comprises an elevated rotating cylinder upon which a rodent is placed (Jones & Roberts, 1968; Pritchett & Mulder, 2003). In order to avoid falling off the rotarod, the mouse must maintain constant motion; hence it is a test of forced motor activity (Jones & Roberts, 1968; Pritchett & Mulder, 2003). To remain on the rotarod while it accelerates at a set rate the mouse requires balance and coordination. Mice were placed on the rotarod and the accelerating rotarod protocol was used. Specifically, the rod starts to rotate at 6 revolutions per minute (rpm) and is then increased in 4 rpm increments up to a maximum of 50 rpm. The experiment continues until the mouse falls off, or until the cut-off time of 300 seconds has elapsed.\n\nThermal nociception. A modified tail flick test (D’Armour & Smith, 1941; Mogil, 2009) was employed as follows: Thermal nociceptive thresholds were assessed by the immersion of 2cm of the animal’s tail in a water bath maintained at a specified temperature such as 40–50°C until the tail flick manoeuvre was initiated. The tail flick latency time was recorded and this parameter was used to compare the effect of neurosteroids versus injection of control vehicle and baseline measurements. A cut off time of 15 seconds was used to minimise the likelihood of tissue damage. The results of each experiment may be expressed in seconds, or as a percentage of the maximum possible effect (MPE) i.e. a percentage of 15 seconds.\n\nMechanical nociception. A series of calibrated von Frey filaments (Ugo Basile, It) were employed to characterise mechanical nociceptive thresholds in WT and ob/ob mice. The mice were placed into clear plastic cubicles on top of a raised platform (Ugo Basile, Italy) with a meshed surface and allowed to acclimatise to the new environment for 30 minutes. The tip of the von Frey filament was pressed carefully onto the middle of the ventral (‘palmar’) surface of the hindpaw. Sufficient force to induce bending of the shaft of the filament was applied for up to 5 seconds and the presence of a withdrawal response noted if it occurred. The procedure was repeated until each hindpaw had received five presses of the filament. Only robust and immediate withdrawal responses were considered as positive. Each mouse could have a maximum score of 10 (five for each hindpaw). Testing would commence with the thinnest filament used and then progress to thicker filaments. Pilot studies were carried out to determine the optimal four filaments to be used in the mice: 0.16g, 0.4g, 0.6g and 1g filaments. They elicited a response in approximately 20%, 40%, 60% and 90% of occasions when applied to adult WT mice. This meant that these filaments could be used to test for the presence of mechanical hypersensitivty and nociception in the ob/ob mouse. This method was adapted from work published in rats with neuropathic sensitisation (Meyer et al., 2011).\n\nStatistical analysis. All data in the results section are expressed as the arithmetic mean ± the standard deviation (SD). The following statistical tests were employed where appropriate for the electrophysiological data: Student’s t test (Excel, Microsoft Office), One-way ANOVA, One-way and Two-way RM ANOVA (Sigmastat). The following non-parametric statistical tests were employed for the behavioural data: Mann-Whitney Rank Sum test, Kruskal Wallis one-way ANOVA on ranks and the Wilcoxon signed rank test (before & after; Sigmastat).\n\n\nResults\n\nElectrophysiological recordings were made from C57/Bl6 mice in lamina II (LII) of the dorsal horn of the spinal cord, the nucleus reticularis (nRT) of the thalamus and layer 2/3 of the cortex (pyramidal neurons). These neurons were selected due to their modulatory role in nociceptive transmission, associated with the perception of pain in humans (Clasca et al., 2012; Cox et al., 1997; DeFelipe & Farinas, 1992; Gentet & Ulrich, 2003). Previous work (Brown et al., 2015) on murine neurons of the cortex and thalamocortical neurons of the ventrobasal (VB) thalamus revealed that fluctuations in the endogenous neurosteroid tone during development (P7–P24) influenced the duration of miniature inhibitory postsynaptic currents (mIPSCs).\n\nPrevious studies of LII neurons revealed the mIPSC time course of decay may reduce with development, a perturbation that may be caused by the loss of an endogenous steroid tone (Keller et al., 2001; Keller et al., 2004; Maguire et al., 2014; Rajalu et al., 2009). mIPSC decay (τW) decreased with development (P8–11 = 24.8 ± 2 ms; n = 26; P17–25 = 19.4 ± 1.8 ms, n = 31; P60–75 = 17.5 ± 1.8 ms, n = 13; One-way ANOVA, P <0.05; Figure 2A,B). The τW is approximately equivalent to the time taken for the mIPSC to decrease by 67% from the peak amplitude. The unfavourable signal-to-noise of individual mIPSCs precludes the accurate fitting of the τW to individual mIPSCs. Therefore, this function is fitted to the mean mIPSC derived for each neuron.\n\n(A) Superimposed exemplar GABAAR-mediated mIPSCs recorded from representative spinal neurons from three stages of development: P8–11 (light grey), P17–25 (grey) and P60–75 (black). Note the reduction of GABAAR mIPSC decay time that occurs with development. (B) Histogram illustrating the shortening of GABAAR mIPSCs with development (One-way ANOVA *P < 0.05; n = 13–31). (C & D) A parallel developmental change is observed in nRT neurons One-way ANOVA P < 0.05. Post hoc Newman Keul’s test revealed significant differences between all groups, **P < 0.05; n = 14–24). (E & F) A parallel developmental change is observed in L2/3 cortical neurons One-way ANOVA P < 0.05. Post hoc Newman Keul’s test revealed significant differences between all groups, **P < 0.05; n = 7–35).\n\nThe nRT is the main source of GABAergic input into the thalamus and there are reciprocal GABAergic loops of innervation between nRT neurons and those of the VB. The two types of neurons, nRT and VB; regulate each other’s function by this mutual inhibitory mechanism (Arcelli et al., 1997; Cox et al., 1997; Gentet & Ulrich, 2003; Guillery & Harting, 2003). The inter-relationship between the nRT and VB acts to modulate nociceptive transmission (Huh et al., 2012). Recordings from nRT neurons were made at three developmental stages: P6-7, P9-10 and P17-25. It is not practical to make such recordings from mice over the age of P25 due to the high density of axonal projections (Cox et al., 1997; Pinault & Deschenes, 1998). The mIPSC decay (τW) of nRT neurons decreased significantly with development (P6-7 = 33.2 ± 1 ms, n = 24; P9-10 = 22.5 ± 0.7 ms, n = 14; P17-25 = 18.2 ± 0.6 ms, n = 32; One-way ANOVA, P < 0.05; post hoc Newman Keul’s test revealed significant differences between all groups, P <0.05; Figure 2C,D).\n\nFurther experiments revealed that GABAARs from neurons at all three levels of the pain pathway are sensitive to modulation by neurosteroids (see Additional material). The endogenous neurosteroid tone was explored using γ-cyclodextrin (γ-CD), a barrel-shaped molecule known to sequester neurosteroids (Shu et al., 2004; Shu et al., 2007). The three principle types of CD are the α-CD, β-CD and γ-CD, they have internal diameters of 5.2 nm, 6.4 nm and 8.3 nm respectively (Cooper et al., 2005; Davis & Brewster, 2004). The largest of these, γ-CD is the most effective for the sequestration of neurosteroids (Brown et al., 2015; Shu et al., 2004; Shu et al., 2007). The mIPSC decay (τW) in WT mice was significantly decreased in the presence of γ-CD, but not by α-CD, or β-CD (see Additional material). γ-CD has been reported to have no direct effect on the GABAAR (Shu et al., 2004; Shu et al., 2007) and neither of the smaller molecules, α-CD or β-CD, had an impact on mIPSC τW (Table S2). Intracellular application of γ-CD via the recording pipette was found to be the optimal method of application (see Additional material) and is consistent with the hypothesis that the GABAAR-active neurosteroids are synthesised within the pain pathway neurons themselves (Akk et al., 2005; Chisari et al., 2009; Tsutsui, 2008).\n\nThe mIPSC decay (τW) of L2/3 cortical neurons at two stages of maturity (P9–10 and P60–75) was significantly decreased in the presence of intracellular γ-CD (P <0.05, Table S1). These data are in contrast to the lack of effect of γ-CD observed in nRT neurons at P9–10 and P17–24. However, the data for P9/10 L2/3 cortical neurons are consistent with data published previously (Brown et al., 2015). Different regions of the nervous system reach maturation at different ages and it is possible that this may account for the regional variations observed. Interestingly, the endogenous neurosteroid tone that previously appeared to be lost during maturation re-emerges in the adult mouse cortex, which may have a significant physiological role.\n\nAs the behavioural studies were to be conducted in adult mice, it was decided to make recordings from adult layer 2/3 cortex neurons for three reasons: 1) Viable recordings of nRT neurons of older animals are compromised by the high density of axonal projections (Cox et al., 1997; Pinault & Deschenes, 1998). 2) Values for the mean τW of L2/3 cortical GABAAR mIPSCs are relatively homogenous, in contrast to the mean τW values of GABAAR mIPSCs of LII neurons which are heterogenous (Mitchell et al., 2007), which makes inter-group comparison more difficult. 3) Layer 2/3 pyramidal neurons of the somatosensory cortex are part of the pain pathway.\n\nThe developmental age of P60–75 was chosen because it facilitated a comparison with the ob/ob mouse model of type-2 diabetes mellitus (T2DM). The ob/ob mouse develops super-morbid obesity and exhibits a neuropathic phenotype and consequently develops hypersensitivity to pain by the age of P60–75 (Drel et al., 2006; Latham et al., 2009). To date, there are no published reports of the electrophysiological characterisation of GABAAR function for the ob/ob mouse.\n\nMice were weighed in order to confirm the presence of obesity. The ob/ob and db/db mice (P60–75) both had significantly greater body weights than the respective WT animals of the same age (Table S3, P < 0.05). These data are consistent with the literature (Bates et al., 2005; Latham et al., 2009).\n\nTo exclude the presence of a direct effect of leptin itself recordings were additionally made in the db/db, which is able to synthesise leptin but lacks the receptor. There was a modest but significant reduction in the mIPSC decay (τW) of cortical neurons between the diabetic mice and the corresponding WT littermates, but no significant difference in the τW, between the three WT strains (P <0.05; Table S4, Figure 3A,C).\n\nRecordings were made to determine that neither the strain, nor the lack of leptin (directly) was an important factor in the shortening of mIPSC τW with γ-CD. In the presence of intracellular γ-CD, there was no significant difference in the mIPSC τW of between all five types of mice (Table S5, Figure 3D, P > 0.05). These findings suggest that synaptic GABAAR function is very similar across all these strains of mice included in the study when the endogenous neurosteroid tone is removed by γ-CD. The results are also consistent with the hypothesis that there is a neurosteroid tone at P60–75, but that it is reduced in both mouse models of diabetic neuropathy. However, the data do not exclude the possibility that the sensitivity of L2/3 cortical GABAARs to neurosteroids may be reduced.\n\n(A) Superimposed exemplar GABAARs mIPSCs from three WT strains (black) and two diabetic phenotypes (grey). (B) Superimposed exemplar GABAAR-mediated mIPSCs from three WT strains (black) and two diabetic phenotypes (grey) with 0.5 μM γ-CD administered intracellularly. (C) Histogram illustrating the shorter cortical GABAAR mIPSC τw of the diabetic mice and their corresponding WT strains (n = 8–25; Student’s unpaired t test *P <0.05). (D) Histogram illustrating that there was no significant difference between all five strains of mice in the presence of intracellular (γ-CD n = 5–15; One-way ANOVA P > 0.05).\n\nThe lipophilic intravenous anaesthetics etomidate and propofol, which in common with neuroactive steroids enhance the function of GABAARs, require relatively prolonged incubation times to approach equilibrium in a brain slice preparation- over 1–2 hours (Benkwitz et al., 2007; Gredell et al., 2004). It is conceivable that the same is true for neurosteroids (Li et al., 2007) therefore recordings were made after incubation treatment with allopregnanolone and ganaxolone. In contrast to the relatively modest prolongation of mIPSCs described above with an acute steroid application protocol (τW: control = 4.0 ± 0.3 ms, n = 7; allopregnanolone 1 μM = 4.5 ± 0.4 ms, n = 7; paired Student’s t test, P <0.05), a 2-hour incubation of the brain slice preparation with nanomolar concentrations of allopregnanolone (100 – 300 nM) produced a dramatic concentration-dependent increase of the WT GABAAR mIPSCs (Table S6, Figure 4A,C,D; P <0.05). These results indicate that allopregnanolone is a potent modulator of synaptic GABAARs in mature cortical neurons, but additionally demonstrate that the steroid effect is greatly underestimated when applied acutely. The large difference between acute bath application and the 2-hour incubation (see Additional material) is probably a consequence of the time required for the steroid to approach equilibrium within the brain slice. No such effect is observed in time-matched controls.\n\nA 2-hour incubation of the ob/ob brain slice preparation with allopregnanolone produced a clear concentration-dependent prolongation of GABAAR mIPSC τW (Table S6, Figure 4B–D; P <0.05). The effect of allopregnanolone was similar for WT, ob/ob and db/db mice (Table S6; P > 0.05). When the data were normalised to reflect the control mIPSC τW, there was no significant difference in the effect of between the three types of mice (Figure S1, P > 0.05). These results suggest that the sensitivity of cortical GABAARs to allopregnanolone in diabetic mice is similar to age-matched WT.\n\nA 2-hour incubation with ganaxolone produced a significant concentration-dependent increase of the WT mIPSC τW, although the magnitude of the effect was less than that induced by allopregnanolone (Table S7, Figure 4E,G,H; P <0.05). In the ob/ob and db/db brain slices incubated with ganaxolone the mIPSC τW prolongation was comparatively greater than for equivalent WT neurons, but there was no significant difference between the ob/ob and db/db (Table S7; P > 0.05). These observations contrast with that of allopregnanolone. These findings suggest that there is a difference in the effect of ganaxolone incubation treatment (but not allopregnanolone) in the ob/ob and db/db mice compared to the WT mice.\n\n(A & B) Superimposed exemplar GABAAR mIPSCs from representative control cortical neurons from mature WT and ob/ob mice respectively and from equivalent neurons after ~2 hour brain slice incubation with 100 nM and 300 nM allopregnanolone. (C) Histogram illustrating the concentration-dependent effect of allopregnanolone on the cortical GABAAR mIPSCs of ob/ob mice (n = 6–25; One-way ANOVA P <0.05. Post hoc Newman Keuls test revealed significant differences between control and both concentrations of allopregnanolone, which increased τw to 362 ± 27% and 534 ± 28% of control respectively *P <0.05). Note that there is no significant difference in response between the two types of mice (n = 8–9; One-way RM ANOVA P > 0.05). (D) Histogram comparing the concentration-dependent effects of allopregnanolone on the duration of GABAARs mIPSC τw expressed as a percentage of control for WT (black) and ob/ob (grey) mice. The histogram illustrates that there is no significant difference in the effect of 300 nM allopregnanolone on the cortical GABAARs mIPSCs of WT and ob/ob neurons (n = 8–9; One-way RM ANOVA P > 0.05). (E & F) Superimposed exemplar GABAAR mIPSCs from representative control cortical neurons from mature WT and ob/ob mice respectively and from equivalent neurons after ~2 hour brain slice incubation with 30 nM - 1 μM ganaxolone. (G) Histogram illustrating the concentration-dependent effect of ganaxolone 30 nM - 1 μM on the GABAARs mIPSCs of ob/ob mice (grey bars; One-way ANOVA P <0.05). Note the exaggerated effect of ganaxolone incubation treatment on the cortical GABAARs mIPSCs of ob/ob mice (grey bars) in comparison to WT mice (black bars) for 30 nM and 300 nM ganaxolone (One-way RM ANOVA, *P < 0.05 respectively). (H) Histogram comparing the concentration-dependent effects of ganaxolone on the duration of GABAARs mIPSC τw expressed as a percentage of control for WT (black) and ob/ob (grey) mice. The histogram illustrates that ganaxolone has an exaggerated effect in ob/ob mice One-way RM ANOVA **P < 0.05). Ctrl = control; Allo = allopregnanolone; Ganax = ganaxolone.\n\nProgesterone and its metabolite dihydroxy-progesterone (DHP) do not modulate GABAARs directly (Belelli & Herd, 2003; Brown et al., 2015), but require the activity of the enzymes 5α-R and 3α-HSD in order to synthesise allopregnanolone (Figure 1; Schumacher et al., 2012; Stoffel-Wagner, 2003).\n\nA 2-hour incubation with progesterone produced a relatively modest prolongation of mIPSC τW in WT mice. The highest concentration of progesterone tested was only slightly more effective than the lowest concentration investigated here (Table S8, Figure 5A,C; P <0.05). These results suggest that the enzymatic function (5α-R and 3α-HSD) is intact and neurosteroids may be synthesised with brain slice incubation of the precursor.\n\n(A & B) Superimposed exemplar GABAAR mIPSCs from a representative control mature WT and ob/ob cortical neurons and from equivalent neurons after ~2 hour brain slice incubation with 1 μM – 50 μM progesterone. (C) Histogram illustrating the significant but modest effect of progesterone incubation treatment on WT and ob/ob cortical GABAARs mIPSCs (grey) that was not concentration-dependent (One-way ANOVA *P <0.05). (D) Histogram illustrating the concentration-independent effect of progesterone on the duration of GABAARs mIPSC τw expressed as a percentage of control. Note that when the effect of progesterone 50 μM was expressed as a percentage of the representative control, the steroid had a greater impact on the ob/ob mice (n = 7–12, One-way RM ANOVA **P < 0.05). (E & F) Superimposed exemplar GABAAR mIPSCs from a representative control mature WT and ob/ob cortical neurons and from equivalent neurons after ~2 hour brain slice incubation with 1 – 3 μM DHP. (G) Histogram illustrating the significant concentration-dependent effect of DHP on the duration of WT and ob/ob cortical GABAARs mIPSCs (n = 10–25; One-way ANOVA P <0.05). Note the exaggerated response of the ob/ob vs. the WT cortical neurons for 1–3 μM DHP (n = 9–10, one-way ANOVA *P < 0.05). (H) Histogram illustrating that 1 – 3 μM DHP has an exaggerated effect in ob/ob (grey) vs. WT (black) mice when the effect is expressed as a percentage of the respective control value (One-way RM ANOVA **P < 0.05). Ctrl = control; DHP = dihydroxyprogesterone; Prog = progesterone.\n\nProgesterone incubation treatment also produced a similarly modest prolongation of mIPSC τW in the ob/ob and db/db (Table S8, Figure 5B,C; P <0.05). When the data for τW were normalised. When the effects of 50 μM progesterone were expressed as a percentage of the representative control, the steroid had a greater impact on the diabetic mice (P < 0.05), but there was no intergroup difference between the ob/ob and db/db mice (P > 0.05; Table S8, Figure S1; Figure 5D).\n\nBath application of DHP (3μM) had no effect on the GABAAR mIPSCs in WT neurons (τW: Control = 5 ± 0.3 ms, n = 4; 3 μM DHP = 4.7 ± 0.3 ms, n = 4, P > 0.05). These findings are consistent with previous work on VB neurons (Brown et al., 2015). The lack of effect after the acute application of DHP contrasts to the modest effects of acutely applied allopregnanolone and ganaxolone (see Additional material). Contrastingly, 2 hours of incubation with DHP (1 – 3 µM) produced a significant, concentration-dependent prolongation of GABAAR mIPSCs in WT (P <0.05; Table S9, Figure 5E,G). These results indicate that mature WT L2/3 cortical neurons have intact 3α-HSD enzymatic function and are able to convert DHP into the active metabolite allopregnanolone. 3μM DHP incubation produced a more pronounced effect in the ob/ob and db/db mice compared to the WT (P <0.05, Table S9), indicating that not only is 3α-HSD enzymatic function preserved in mature mice with T2DM, but may be up-regulated.\n\nFinasteride itself has no direct effect on GABAAR mIPSCs, but pre-treatment with this 5α-R enzyme inhibitor prevents the conversion of progesterone into GABAAR-active neurosteroid (Sanna et al., 2004). Recordings were made after at least 2 hours of incubation with finasteride and progesterone. Finasteride alone had no effect on WT mIPSC τW but it did prevent the effect of progesterone (Table S10; P <0.05). Finasteride alone also had no effect on ob/ob mIPSC τW of mice, but it did prevent the effect of progesterone (Table S10; P < 0.05) indicating that progesterone requires the 5α-R for it to be converted to its’ active metabolites, (although do not prove neurosteroid synthesis by the neurons itself). Finasteride alone had no effect on mIPSC τW, despite the suggested presence of a modest endogenous neurosteroid tone in mature WT mice. This apparent paradox may be explained by comparing how finasteride and γ-CD act. γ-CD will remove the endogenous neurosteroid present, whereas although finasteride should prevent new allopregnanolone synthesis, it will have little impact on that already present. Therefore, the apparent lack of an effect of finasteride may reflect the relatively slow turnover of pre-synthesised allopregnanolone and is consistent with the literature (Brown et al., 2015).\n\nRecordings were made in WT mice after at least 2 hours of incubation with provera and DHP, or provera and ganaxolone. Provera alone exerted a modest effect on mIPSC τW but, in addition, it prevented the effect of DHP (Table S11, Figure S2 A,B; P <0.05). Provera did not inhibit the effect of ganaxolone on τW in WT mice (Table S11, Figure S2 C,D, P >0.05). These results confirm that DHP requires enzymatic conversion by 3α-HSD for it to induce a prolongation of τW. In contrast, ganaxolone was unaffected by 3α-HSD inhibition with provera.\n\nIndometacin alone had no effect on τW, but it also prevented the effect of DHP (Table S12, Figure S2 E,F; P <0.05). By contrast, indometacin had no effect on the prolongation of mIPSC τW by pre-incubated allopregnanolone (Table S13, Figure S2 G,H; P >0.05) confirming that allopregnanolone does not require 3α-HSD in order to modulate GABAAR mIPSCs. Indometacin did not increase the effect of allopregnanolone by potentially inhibiting its metabolism to an inactive form. These experiments with another 3α-HSD enzyme inhibitor confirm that DHP requires conversion to a more active form by 3α-HSD in order for it to induce prolongation of GABAAR mIPSC decay time. By contrast, indometacin exhibited concentration-dependent inhibition of ganaxolone incubation treatment on τW (Table S13, Figure S2 I,J, P <0.05). These results with ganaxolone were unexpected given the lack of impact that indometacin had on the effectiveness of allopregnanolone. Indeed, inhibiting 3α-HSD with provera had no impact on ganaxolone, but prevented the effect of DHP on GABAAR mIPSC τW. This raises the question as to whether ganaxolone is an allosteric modulator of the GABAAR and a precursor to a more active neurosteroid such as allopregnanolone. Alternatively, it raises the question of whether indometacin could be a silent competitive steroid antagonist at the GABAAR and prevent the action of ganaxolone by that mechanism (explored in next section).\n\nAcute allopregnanolone had only a modest effect on the GABAAR-mediated mIPSCs decay time (see Additional material), but was far more efficacious in this respect when brain tissue slices were incubated with steroid for > 2 hours, suggesting that the steroid is relatively slow to equilibrate within the tissue. It has been proposed that endogenous neurosteroids may be synthesised in the postsynaptic neuron and act in an autocrine manner to influence GABA-ergic transmission (Agis-Balboa et al., 2006; Lambert et al., 2009). It is implicit in this model that intracellular steroid would modulate the GABAARs of the postsynaptic neuron and the recording pipette can be employed to deliver drugs to the neuron interior (Evans & Marty, 1986). Recordings were made with allopregnanolone, or ganaxolone present in the recording pipette to explore whether the steroid could modulate synaptic GABAARs when delivered to the intracellular compartment. These recordings were compared to separate control recordings (i.e. they were not paired). The presence of allopregnanolone in the recording pipette significantly increased the WT mIPSC τW in a concentration-dependent manner (Table S14, Figure S3 A,C; P < 0.05). Here, the higher concentration of allopregnanolone (10 µM) had a comparatively large concentration-dependent effect on GABAAR mIPSCs when presented acutely within the recording electrode. This finding indicates that neurosteroids are able to exert their effect via the intracellular compartment and is in agreement with the literature (Akk et al., 2005).\n\nThe concentrations of pipette-applied allopregnanolone are relatively high, but the time scale from application to recording is short (< 10 minutes) and dialysis rate of the intracellular contents may be an influential limiting factor. The presence of ganaxolone in the recording pipette also significantly increased the WT mIPSC τW in a concentration-dependent manner (Table S15, Figure S3 E,G, P < 0.05). The effect of pipette-applied ganaxolone is consistent with the recordings described above with allopregnanolone, although ganaxolone had a less pronounced effect on GABAAR mIPSCs. Collectively these experiments illustrate that steroid penetration of brain slice tissue is a significant limiting factor, the intracellular application is an effective method of presenting neuroactive steroids to GABAARs (Table S14–Table S16, Figure 1, Figure S3 A–H) and that, in these neurons, allopregnanolone is a more effective modulator of the synaptic GABAAR than ganaxolone (Figure S3 C,D,G,H).\n\nRecordings were also made in ob/ob neurons with allopregnanolone, or ganaxolone, present in the recording pipette. The ob/ob mIPSC τW was increased by the presence of allopregnanolone in the recording pipette (Table S14, Figure S3 B,C; P < 0.05). After normalisation, intracellular allopregnanolone 3 – 10 μM increased τW to similar percentages of the strain representative controls (Table S14, Figure S3 D, P > 0.05). These findings are consistent with the hypothesis that there is no difference in the sensitivity of the cortical GABAAR to neurosteroids between the ob/ob and WT mice.\n\nThe ob/ob mIPSC τW was increased by the presence of ganaxolone in the recording pipette (Table S15, Figure S3 F,G, P < 0.05). Ganaxolone did not have a concentration-dependent effect on ob/ob cortical GABAAR mIPSCs when acutely delivered intracellularly and the magnitude of the effect was less than that produced by equivalent pipette concentrations of allopregnanolone (Table S14–Table S15, Figure S3 C,D,G,H,L, P > 0.05). There was no difference in the GABAAR mIPSC τW between the ob/ob and the WT when treated with intracellular ganaxolone. This finding is consistent with the recordings described above with allopregnanolone, although intracellular ganaxolone had a less pronounced effect.\n\nIn order to explore whether indometacin could be a silent competitive antagonist at the GABAAR (and therefore prevent modulation by ganaxolone) recordings were made with indometacin ± ganaxolone presented in the pipette. The WT mIPSC τW was unchanged by the presence of indometacin in the pipette and indometacin had no effect on the modulatory action of ganaxolone (Table S16, Figure S3 I,J, P >0.05).\n\nRecordings were also made in ob/ob neurons with indometacin ± ganaxolone presented in the pipette. The mIPSC τW in ob/ob mice was unaffected by the presence of indometacin in the pipette and indometacin had no effect on the modulatory action of ganaxolone (Table S16, P >0.05). It is possible that indometacin may inhibit the action of ganaxolone by preventing its conversion to a more active compound (such as allopregnanolone).\n\nThe WT and ob/ob mIPSC τW was unaffected by the presence of DHP in the recording pipette (Table S17, P > 0.05). These findings contrast with the significant effects of intracellular allopregnanolone and ganaxolone and are consistent with the idea that DHP is a precursor compound (Figure 1, Figure S3 K,L). Considering these results it would be difficult to make the case for the alternative hypothesis that GABAAR sensitivity is increased in ob/ob and db/db mice.\n\nThe developmental age of P60–75 was chosen for both electrophysiological and behavioural experiments as it represents physiological maturity and facilitates a comparison with the ob/ob mouse model of T2DM, which develops hypersensitivity to pain by this age (Drel et al., 2006; Latham et al., 2009).\n\nThe ob/ob mice were able to remain on the rotarod for a significantly shorter time than were WT mice (WT = 216 ± 14 s, n = 15; ob/ob = 21 ± 5 s, n = 15; Mann-Whitney Rank Sum test, P < 0.001; Figure 6A). The impaired rotarod performance of the ob/ob mouse is consistent with a previous report (Mayers et al., 2009). It is not possible to have obese WT mice as controls because those obese mice would be likely to have similar neurological consequences as the ob/ob mice.\n\nGanaxolone impairs thermal nociception in WT mice at doses low enough to preserve rotarod performance. Ganaxolone reduces mechanical hypersensitivity in ob/ob mice and reduces mechanical nociceptive pain in both WT and ob/ob mice. (A) Histogram illustrating the dramatic impairment of sensorimotor function exhibited by mature ob/ob mice. The ob/ob mice were able to remain on the accelerating rotarod for a significantly shorter time than were WT mice (n = 15; Mann-Whitney Rank Sum test P < 0.001). (B) Histogram illustrating that mature ob/ob mice have a delayed response to thermal noxious stimuli in comparison to age-matched WT mice. (P < 0.05). There was no significant difference at the less noxious temperature of 46°C (n = 20 per group, Mann-Whitney Rank Sum test P > 0.05). A maximum withdrawal latency of 15 seconds was enforced to minimise the possibility of tissue damage. (C) Histogram illustrating that ob/ob mice exhibit mechanical hypersensitivity in comparison to WT mice. The ob/ob mice responded significantly more frequently than the WT mice to the 0.16 g, 0.4 g and 0.6 g filaments (P < 0.05). By contrast, there was no significant difference in the frequency of responses to the 1 g vF filament (Mann-Whitney Rank Sum test P > 0.05). These results confirm the presence of mechanical hypersensitivity in the ob/ob mouse. (D) Histogram illustrating that ganaxolone induced a dose-dependent prolongation of tail withdrawal latency in WT mice after 30 minutes (P < 0.05). The solubilising vehicle β-CD had no effect (Kruskall Wallis One-Way ANOVA on ranks P > 0.05). (E) Histogram showing that at 15, 30 and 60 minutes post-injection the highest dose of ganaxolone (30mg/kg) significantly impaired performance of WT mice on the rotarod (P < 0.05). The effect of ganaxolone (30 mg/kg) was no longer apparent at 120 minutes post-injection (P > 0.05). The solubilising vehicle β-CD had no effect on rotarod performance (Kruskall Wallis One-Way ANOVA on ranks P > 0.05). (F) Histogram illustrating that ganaxolone had no impact on the response to the 0.16 g, 0.4 g or 0.6 g vF filaments in WT mice (P > 0.05) but ganaxolone did reduce the response of WT mice to the 1.0 g vF filament (n = 10 per group, Wilcoxon signed rank test (before & after) P < 0.05). These data suggest that ganaxolone is analgesic for mechanical nociceptive pain in WT mice. WT mice do not exhibit hypersensitivity under normal conditions; therefore it is not unexpected that these drugs had no effect on the response to the smaller vF filaments. (G) Histogram illustrating that ganaxolone had no impact on the response to the 0.16 g vF filament the ob/ob mouse (P > 0.05). In contrast, ganaxolone reduced the response of the ob/ob mouse to the 0.4 g, 0.6 g and 1 g vF filaments (n = 10 per group, Wilcoxon signed rank test (before & after) P < 0.05). Ganaxolone reduced the response of the ob/ob mouse to the 0.4 g and the 0.6g vF filaments, which is consistent with the idea that this neuroactive steroid may reduce mechanical hypersensitivity. As described earlier, the 1.0 g vF filament induces a withdrawal response on ~85% of occasions in the WT and ~93% in the ob/ob mouse and may therefore be considered as an unambiguous test of mechanical nociceptive pain. Taken as a whole, these results show that ganaxolone was effective for mechanical nociceptive pain in both strains of mice.\n\nThe tail flick test was used to characterise the response of the WT and the ob/ob mouse to three distinct temperatures. At 48°C and 50°C ob/ob mice exhibited significantly longer tail withdrawal latencies compared to WT mice (Table S18, Figure 6B, P < 0.05). By contrast, there was no significant difference with a water temperature of 46°C (Table S18, Figure 6B, P > 0.05). These findings are consistent with the phenomenon of thermal hypoalgesia in ob/ob mice reported by Drel et al. (2006), but conflict with the thermal hyperalgesia reported by Latham et al. (2009).\n\nThe tail flick test is a useful tool for assessing thermal nociception. Ganaxolone was considered a priori considered to be a metabolically stable synthetic analogue of allopregnanolone, the molecular structure of which differs only by having an extra methyl group that prevents oxidation to an inactive form and has previously been used in clinical trials (Carter et al., 1997). Ganaxolone required to be solubilised with hydroxypropyl β-CD prior to intraperitoneal injection (Besheer et al., 2010; Carter et al., 1997; Reddy & Rogawski, 2010). There were no significant differences in the baseline withdrawal latencies in response to a noxious thermal stimulus (50°C) between the four groups of mice used to examine the effects of ganaxolone (Table S19, Figure 6C, P > 0.05). Ganaxolone induced a dose-dependent prolongation of tail withdrawal latency in WT mice after 30 minutes and this effect lasted more than 90 minutes (Table S19, Figure 6C, P < 0.05). These data suggest that ganaxolone exhibits a dose-dependent analgesic effect in a test of thermal nociception in WT mice. This finding is consistent with the intrathecal administration of ganaxolone (Asiedu et al., 2012) and also other reports of the analgesic effects of similar neurosteroids in rats in the setting of post-chemotherapy neuropathy (Meyer et al., 2010; Meyer et al., 2011).\n\nOnly the highest dose of ganaxolone significantly impaired the performance of WT mice on the rotarod. There was no significant difference in the baseline rotarod performance of the four groups in the study (Table S20, Figure 6D, P > 0.05). However, at 15, 30 and 60 minutes post-injection the highest dose of ganaxolone (30 mg/kg) significantly impaired performance of WT mice on the rotarod (Table S20, Figure 6D, P < 0.05). The effect of ganaxolone (30 mg/kg) was no longer apparent at 120 minutes post-injection (Table S20, Figure 6D, P > 0.05). Taken in conjunction with the tail flick data, these data suggest that ganaxolone exhibits an analgesic effect at a dose of 10 mg/kg but only impairs rotarod performance in WT mice at higher doses such as 30 mg/kg. In addition, the dose-dependent effect on rotarod performance is consistent with the literature (Carter et al., 1997).\n\nvon Frey (vF) filaments have previously been employed via the up-down method to demonstrate that the ob/ob mechanical hypersensitivity by the age of P60 (Drel et al., 2006; Latham et al., 2009). The 0.16 g, 0.4 g and 0.6 g vF filaments elicit withdrawal responses on 25%, 40% and 60% of occasions respectively in WT mice. If the ob/ob mouse has an exaggerated response to these filaments this would be consistent with mechanical hypersensitivity (Merskey & Bogduk, 1994). The 1 g vF filament elicited a withdrawal response on approximately 90% of occasions, therefore it is considered to be a clear test of mechanical nociceptive pain in the WT mouse. This method was adapted from work published in rats with neuropathic sensitisation (Meyer et al., 2011). The ob/ob mice responded significantly more frequently than the WT mice to the 0.16 g, 0.4 g and 0.6 g filaments (Table S21, Figure 6E, P < 0.05). By contrast, there was no significant difference in the frequency of responses to the 1 g vF filament (Table S21, Figure 6E, P > 0.05). These results confirm the presence of mechanical hypersensitivity in the ob/ob mouse, which is consistent with reports in the literature (Drel et al., 2006; Latham et al., 2009).\n\nvF filaments were used for the comparison of mechanical sensitivity of WT and ob/ob mice before and after intraperitoneal drug administration. The β-CD vehicle had no impact on the response to any of the vF filaments in WT or ob/ob mice (Table S22, P > 0.05). Ganaxolone (10 mg/kg) had no impact on the response to the 0.16 g, 0.4 g or 0.6 g vF filaments in the WT mouse (Table S23, Figure 6F, P > 0.05) but ganaxolone did reduce the response of WT mice to the 1.0 g vF filament (Table S23, Figure 6F, P < 0.05). These data suggest that ganaxolone is analgesic for mechanical nociceptive pain in WT mice. By definition, WT mice do not exhibit hypersensitivity or allodynia under normal conditions; therefore it is perhaps unsurprising that these drugs did not impact on the response to the smaller vF filaments.\n\nGanaxolone had no impact on the response rate to the 0.16 g vF filament in ob/ob mice (Table S24, Figure 6G, P > 0.05), but in contrast, ganaxolone reduced the response rate of ob/ob mice to the 0.4 g, 0.6 g and 1 g vF filaments (Table S24, Figure 6G, P < 0.05). ob/ob mice have exaggerated baseline response rates to the 0.16 g, 0.4 g and 0.6 g vF filaments in comparison to WT mice (Figure 6E). Ganaxolone reduced the response rates of ob/ob mice to the 0.4 g and 0.6 g vF filaments, which is consistent with the idea that these neurosteroids may reduce mechanical hypersensitivity. As described earlier, the 1.0 g vF filament induces a withdrawal response on ~85% of occasions in WT mice and ~93% in ob/ob mice (Figure 6E) and may therefore be considered as an unambiguous test of mechanical nociceptive pain. Taken as a whole, these results show that ganaxolone was effective for mechanical nociceptive pain in both strains of mice.\n\n\nDiscussion\n\nThe decay time of GABAAR mIPSCs decreases with development at three levels of the pain pathway. GABAARs from pain pathway neurons are sensitive to modulation by neurosteroids, and this was explored using γ-CD applied within the recording pipette suggesting that neurosteroids may be synthesised within neurons themselves. The endogenous neurosteroid tone that previously appeared to be lost with maturation re-emerges in the adult mouse cortex, which is suggestive of a significant physiological role. Layer 2/3 pyramidal neurons of the somatosensory cortex are involved in the pain pathway, exhibit relatively homogenous GABAAR mIPSCs and are accessible for electrophysiological experimentation in mature animals. In contrast, there are inherent difficulties in studying neurons from other parts of the pain pathway in mature mice. Specifically, there is a relatively wide variation in decay time of GABAAR mIPSCs from Lamina II neurons of the spinal cord (Mitchell et al., 2007) making it very challenging to assess the effects of pharmacological agents on decay time. Separately, thalamic neurons of mice over the age of P25 have a high density of axonal projections that obscures visualisation of individual neurons (Cox et al., 1997; Pinault & Deschenes, 1998).\n\nThere are no published reports of the electrophysiological characterisation of synaptic GABAAR function for T2DM. It was unknown whether neurosteroid tone would be altered in mice with diabetic neuropathy. In inflammatory pain, neurosteroidogenesis is upregulated to mediate a form of endogenous analgesia by enhancing GABAergic neural inhibition (Poisbeau et al., 2005). The ob/ob model of T2DM is particularly useful for the study of neuropathic pain because it develops a more clinically relevant form of diabetes than other models (Cefalu, 2006; Drel et al., 2006; Latham et al., 2009; Lindstrom, 2007; Sullivan et al., 2007). However, it should be noted that no animal model of diabetes fully replicates the human phenotype. ob/ob mice are deficient of the hormone leptin and it was considered that leptin itself may modulate GABAAR function (Solovyova et al., 2009). Therefore, recordings were made in a second mouse model of T2DM, the db/db mouse, which is able to synthesise leptin but lacks the leptin receptor (Bates et al., 2005; Cefalu, 2006).\n\nGABAAR mIPSCs from both ob/ob and db/db mice had significantly shorter decay times than the three WT strains, making it unlikely that leptin is involved. A decrease in the sensitivity of the GABAAR to allosteric modulators such as neurosteroids would be a possible explanation for the shorter GABAAR mIPSCs. However, reduced GABAAR sensitivity is inconsistent with the exaggerated effect of ganaxolone and DHP in ob/ob and the similar effect of allopregnanolone in both WT and diabetic cortical neurons. Pipette-applied allopregnanolone or ganaxolone induced the same concentration-dependent increase in mIPSC decay time, in ob/ob and WT, in keeping with Akk et al. (2005). A possible explanation may be that in T2DM mice there is a reduction in the endogenous neurosteroid tone, which is caused by a common pathological insult. Reduced neurosteroid tone could result in diminished GABAergic inhibition and subsequently a hypersensitive phenotype. Theoretically, other endogenous compounds that modulate the GABAAR such as endocannabinoids may be affected by a common mechanism in T2DM; however the selective effects of the neurosteroid scavenger γ-CD (in comparison to the lack of effect with α-CD and β-CD) make this less likely. γ-CD reduced the GABAAR mIPSC decay time to similar baseline values for all five lines of mice. These findings are consistent with the hypothesis that mature WT mice possess an endogenous autocrine neurosteroid tone that may fine-tune GABAAR function under physiological conditions but is reduced in ob/ob and db/db mice (in which underlying GABAAR function and sensitivity are otherwise unchanged). Thus, in neuropathic pain associated with diabetes, neurosteroid levels may be reduced, which contrasts to inflammatory pain, where neurosteroid levels may be upregulated (Poisbeau et al., 2005).\n\nWith reference to neurosteroidogenesis in the diabetic mice, incubation treatment with progesterone produced a modest prolongation of GABAAR mIPSC decay time in ob/ob mice that was similar to WT controls and was not concentration-dependent from 1–50 μM. Additional recordings made at the highest progesterone concentration in db/db mice were no different to those of the WT or ob/ob. This suggests that enzymatic function responsible for the endogenous synthesis of allopregnanolone is intact in ob/ob and db/db. Indeed, due to the shorter baseline decay time of mIPSCs of ob/ob and db/db mice, progesterone actually had a proportionately greater effect in the diabetic mice than in WT controls. In keeping with what had been observed in the WT, incubation treatment with finasteride (5α-R inhibitor) also had no effect on ob/ob mIPSCs directly, but was able to block the effect of progesterone in ob/ob mice.\n\nIncubation treatment with DHP (the progesterone metabolite) produced a significant, concentration-dependent prolongation of GABAAR mIPSC decay time that was exaggerated in diabetic mice compared to WT. This finding suggests that not only is 3α-HSD enzymatic function preserved in mature T2DM mice, but it may actually be upregulated. A potential increase in 3α-HSD activity could be a partial compensation for a reduced baseline endogenous neurosteroid tone in diabetic mice. In keeping with what had been observed in WT mice, pipette-applied DHP had no effect on the ob/ob mIPSCs. The lack of effect of DHP is consistent with DHP’s role as a metabolic precursor to the active compound allopregnanolone.\n\nGanaxolone incubation treatment also had a greater impact in ob/ob mice compared with WT, despite the compound being considered previously to be more metabolically stable (Carter et al., 1997). This finding permits the possibility that ganaxolone may potentially be demethylated to allopregnanolone in order to exert maximal effect on cortical GABAARs. The idea of ganaxolone as both an agonist and precursor to allopregnanolone would also be consistent with the relatively larger effect of ganaxolone incubation in ob/ob mice, which may exhibit an upregulation of key enzymes such as CYP2C and 3α-HSD. The inhibitory effect of indometacin on ganaxolone incubation treatment is also consistent with the notion of ganaxolone as an active precursor of allopregnanolone. Indeed, indometacin had no impact on ob/ob cortical neurons when present in the pipette or after 2 hours of incubation treatment. When indometacin and ganaxolone were co-applied within the pipette, indometacin had no impact on the ability of ganaxolone to prolong mIPSCs. The CYP2C subfamily of enzymes is known to metabolise neurosteroids (McFadyen et al., 1998; Miksys & Tyndale, 2002) and cholesterol is metabolised to pregnenolone by CYP450scc (Miksys & Tyndale, 2002; Schumacher et al., 2012). Therefore, one possible explanation for this is that indometacin could compete with ganaxolone for the demethylating effects of CYP2C9, however, it must be noted that indometacin may have other non-specific effects.\n\nThe underlying mechanism responsible for the reduced neurosteroid tone in the diabetic mice is uncertain, but may be due to mitochondrial dysfunction (Chowdhury et al., 2013; Edwards et al., 2008; Fernyhough et al., 2010; Vincent et al., 2010). It has been postulated that hyperglycaemia has several detrimental effects including the excessive donation of electrons to the mitochondrial electron transport chain, which induces an increased production of reactive oxygen species. The increased availability of electrons may lead to the partial reduction of oxygen to neurotoxic superoxide radicals (Chowdhury et al., 2013; Fernyhough et al., 2010; Vincent et al., 2010). Mitochondrial dysfunction could also account for both mechanical hypersensitivity associated with a reduction in mitochondrial-derived neurosteroids and also for thermal hyposensitivity associated with axonal degeneration (Chowdhury et al., 2013; Drel et al., 2006; Latham et al., 2009; Vincent et al., 2010). The mechanisms of mitochondrial dysfunction in diabetic neuropathy has been covered by other authors (Chowdhury et al., 2013; Fernyhough et al., 2010; Vincent et al., 2010). These data are consistent with the possibility that neurosteroidogenic enzyme function may be upregulated in the ob/ob mice.\n\nThe behavioural work was consistent with the known phenotype of the ob/ob (Drel et al., 2006; Mayers et al., 2009) and translated the electrophysiological findings into measurable anti-nociceptive effects. The polyneuropathy phenotype with seemingly paradoxical mechanical hypersensitivity and thermal hyposensitivity is in fact consistent with the progressive diabetic neuropathy observed in the literature (Cefalu, 2006; Drel et al., 2006; Kaplan & Wagner, 2006; Latham et al., 2009). The mechanical hypersensitivity of ob/ob mice and the increased response to ganaxolone reflected the observed electrophysiological findings. Ob/ob mice developed obesity and T2DM on a normal diet, which led to painful neuropathy by the age of P60-75. At this age they exhibited sensorimotor impairment, thermal hypoalgesia, cold allodynia and mechanical allodynia. In WT mice, the neuroactive steroid ganaxolone impaired sensorimotor function at 30 mg/kg, but the lower dose of 10 mg/kg did not, and was analgesic for thermal and mechanical nociceptive pain. The effect of ganaxolone on thermal nociception and sensorimotor impairment could not be tested on ob/ob mice due to their pre-existing deficits. However, ganaxolone significantly reduced mechanical allodynia in ob/ob mice. These results suggest that GABAAR-modulatory neurosteroidal drugs such as ganaxolone may have analgesic properties for nociceptive pain and also neuropathic pain by restoration of the depressed endogenous neurosteroid tone.\n\n\nData and software availability\n\nOpen Science Framework: Dataset of ‘Neurosteroids are reduced in diabetic neuropathy and may be associated with the development of neuropathic pain’, doi: 10.17605/osf.io/bk3tw (Humble, 2016).\n\nAnalysed raw data can be found in the supplementary tables (See Supplementary material).\n\nPlease refer to Methods section for details of standard software used for data analysis.",
"appendix": "Author contributions\n\n\n\nDr Stephen Humble is responsible for this all work, including planning the experiments, performing the experiments and writing the paper. Prof Hales, Lambert and Dr Belelli assisted Dr Humble with regards to planning many of the experiments. However, after discussion it was decided that their contributions merited being listed in the acknowledgements section rather than as co-authors.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by the Wellcome Trust (Grant No. 090667).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nI am indebted to the generous support of the Wellcome Trust and would also like to thank: Prof Hales, Dr Belelli, Dr McCrimmon, Prof Peters, Prof Sillar, Dr Connolly, Dr Miles, Prof Poisbeau, Prof Lambert for their scientific advice, Miss Gallacher, Miss Wright, Dr McLeod, Mr McLeod, Dr Newman, Dr Cooper, Dr Brown, Dr Panetta, Dr Livesey for their technical assistance, Prof Matthews, Dr Moffat, Mr F Kafka, Prof P Anand, Dr P Donatien, Dr R Privitera, Dr Y Yangou, Prof A Dickenson, Dr Platt, Dr Ladas, Dr Jenner, Dr Feynman, Mrs E Humble for their support.\n\n\nSupplementary material\n\nHistogram summarising the relative differences in the potencies of certain neurosteroidal compounds (allopregnanolone, ganaxolone, DHP and progesterone) in three strains of mice (WT, ob/ob and db/db) after ~2 hours brain-slice exposure. When the mIPSC τw were normalised for comparison, allopregnanolone and ganaxolone, which modulate the GABAAR directly, have the greatest potency, while the neurosteroid precursors DHP and progesterone have the least potency. There is no significant difference in the effect of allopregnanolone on the three strains of mice (n = 6–9, P > 0.05). In contrast, the precursor DHP had a greater impact on the diabetic mice n = 8–10; One-way RM ANOVA *P < 0.05. Post hoc Newman Keul’s test revealed significant differences between the WT and both types of diabetic mice, *P < 0.05, but there was no intergroup difference between the ob/ob and db/db mice, P > 0.05). The precursor progesterone, also had a greater impact on the diabetic mice, n = 6–12; One-way RM ANOVA P < 0.05. Post hoc Newman Keul’s test revealed significant differences between the WT and both types of diabetic mice, P < 0.05, but there was no intergroup difference between the ob/ob and db/db mice, P > 0.05). Ctrl = control; Allo = allopregnanolone; Ganax = ganaxolone; DHP = dihydroxyprogesterone; Prog = progesterone.\n\n(A) Superimposed exemplar averaged GABAAR mIPSCs from a representative control cortical neuron and from equivalent neurons after ~2 hour pre-incubation of the brain slice with provera (1 μM), DHP (3 μM), or both. To facilitate comparison of their kinetics, the amplitude of the mIPSCs are normalised to that of the control averaged mIPSC (B) Histogram illustrating that provera prevents the effect of DHP pre-incubation to prolong GABAAR-mediated mIPSCs (τw in ms; one-way ANOVA P <0.05. Post hoc Newman Keul’s test revealed differences for DHP (3 μM) with or without provera (1 μM), *P <0.05, n = 7–9). (C) Superimposed exemplar averaged GABAAR mIPSCs from a representative control cortical neuron and from equivalent neurons after ~2 hour pre-incubation of the brain slice with provera (1 μM), ganaxolone (300 nM), or both. (D) Histogram illustrating that provera did not prevent the effect of ganaxolone on the duration of GABAAR mIPSCs (τw in ms; one-way ANOVA P <0.05. Post hoc Newman Keul’s test revealed no difference for the effects of ganaxolone (300 nM) with, or without provera (1 μM), P >0.05, n = 7–10). (E) Superimposed exemplar averaged GABAAR mIPSCs acquired from a representative control cortical neuron and from equivalent neurons after ~2 hours pre-incubation of the brain slice with indometacin (30 μM), DHP (3 μM), or both. (F) Histogram illustrating that indometacin prevents the effect of DHP on the duration of the GABAAR mIPSCs (τw in ms, one-way ANOVA P <0.05. Post hoc Newman Keul’s test revealed differences for DHP 3 μM with or without indometacin 100 μM, *P <0.05, n = 6–9). (G) Superimposed exemplar averaged GABAAR mIPSCs acquired from a representative control neuron and from equivalent neurons after an ~2 hour pre-incubation of the brain slice with indometacin (100 μM), allopregnanolone (100 nM), or both. (H) Histogram illustrating that indometacin does not prevent the effect of allopregnanolone to prolong the duration of the GABAAR-mediated mIPSC (τw in ms, one-way ANOVA P >0.05. Post hoc Newman Keul’s test revealed no differences for allopregnanolone with or without indometacin, P >0.05, n =7–9). (I) Superimposed exemplar averaged GABAAR mIPSCs acquired from a representative control neuron and from equivalent neurons after ~2 hour incubation of the brain slice with indometacin (100 µM), ganaxolone (300 nM), or both. (J) Histogram illustrating that indometacin in a concentration-dependant manner (30 – 100 µM) prevents the effect of ganaxolone to prolong the GABAAR-mediated mIPSC (τw in ms; one-way ANOVA P < 0.05. Post hoc Newman Keul’s test revealed differences for ganaxolone 300 nM with or without indometacin 30 – 100 μM, **P < 0.05, n = 8–10. Ctrl = control; Allo = allopregnanolone; Ganax = ganaxolone; DHP = dihydroxyprogesterone; Indo = indometacin.\n\n(A & B) Superimposed exemplar GABAAR mIPSCs from a representative control WT and ob/ob cortical neurons and from equivalent neurons with 3 – 10 μM allopregnanolone administered intracellularly. (C) Histogram illustrating the concentration-dependent effect of the intracellular application of allopregnanolone on the duration of GABAARs mIPSC τw in WT and ob/ob cortical neurons (n = 5–25; One-way ANOVA *P < 0.05). (D) Histogram illustrating the concentration-dependent effect of allopregnanolone on the duration of GABAARs mIPSC τw of WT (black) and ob/ob cortical neurons (grey) expressed as a percentage of control. There was no significant difference in the effect of allopregnanolone on WT vs. ob/ob cortical GABAARs mIPSCs (n = 5–7; One-way RM ANOVA P > 0.05). (E & F) Superimposed exemplar GABAARs mIPSCs from a representative control WT and ob/ob cortical neurons and from equivalent neurons with 3–10 μM ganaxolone administered intracellularly. (G) Histogram illustrating the concentration-dependent effect of the intracellular application of ganaxolone on the duration of GABAARs mIPSC τw in WT and ob/ob cortical neurons (n = 6–25; One-way ANOVA *P < 0.05). (H) Histogram illustrating the concentration-dependent effect of ganaxolone on the duration of GABAARs mIPSC τw of WT (black) and ob/ob cortical neurons (grey) expressed as a percentage of control. There was no significant difference in the effect of ganaxolone on WT vs. ob/ob cortical GABAARs mIPSCs (n = 6; One-way RM ANOVA P > 0.05). (I) Superimposed exemplar averaged GABAAR-mediated mIPSCs acquired from a representative control mature L2/3 cortical neuron and from equivalent neurons after the intracellular application of ganaxolone (10 μM), indometacin (100 μM) or both. (J) Histogram illustrating that intracellular indometacin made no impact on the effectiveness of intracellular ganaxolone (Student’s unpaired t tests P > 0.05 for ganaxolone 10 μM vs. ganaxolone 10 μM with indometacin 100 μM, n = 4–6). (K & L) Histograms illustrating the lack of effect of the intracellular application of 3–10 μM DHP (white bars) on the duration of WT and ob/ob cortical GABAARs mIPSC τw (One-way ANOVA P > 0.05) in contrast to the effectiveness of allopregnanolone and ganaxolone. Ctrl = control; Allo = allopregnanolone; Ganax = ganaxolone; DHP = dihydroxyprogesterone.\n\nAdditional material\n\nClick here to access the data.\n\nSupplementary tables\n\nClick here to access the data.\n\n\nReferences\n\nAgis-Balboa RC, Pinna G, Zhubi A, et al.: Characterization of brain neurons that express enzymes mediating neurosteroid biosynthesis. Proc Natl Acad Sci U S A. 2006; 103(39): 14602–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkk G, Shu HJ, Wang C, et al.: Neurosteroid access to the GABAA receptor. J Neurosci. 2005; 25(50): 11605–13. PubMed Abstract | Publisher Full Text\n\nArcelli P, Frassoni C, Regondi MC, et al.: GABAergic neurons in mammalian thalamus: a marker of thalamic complexity? Brain Research Bulletin. 1997; 42(1): 27–37. PubMed Abstract | Publisher Full Text\n\nAsiedu MN, Mejia G, Ossipov MK, et al.: Modulation of spinal GABAergic analgesia by inhibition of chloride extrusion capacity in mice. J Pain. 2012; 13(6): 546–554. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBates SH, Kulkarni RN, Seifert M, et al.: Roles for leptin receptor/STAT3-dependent and -independent signals in the regulation of glucose homeostasis. Cell Metab. 2005; 1(3): 169–73. PubMed Abstract | Publisher Full Text\n\nBelelli D, Herd MB: The contraceptive agent Provera enhances GABAA receptor-mediated inhibitory neurotransmission in the rat hippocampus: evidence for endogenous neurosteroids? J Neurosci. 2003; 23(31): 10013–20. PubMed Abstract\n\nBenkwitz C, Liao M, Laster MJ, et al.: Determination of the EC50 amnesic concentration of etomidate and its diffusion profile in brain tissue: implications for in vitro studies. Anesthesiology. 2007; 106(1): 114–23. PubMed Abstract | Publisher Full Text\n\nBesheer J, Linsay TG, O'Buckley TK, et al.: Pregnenolone and ganaxolone reduce operant ethanol self-administration in alcohol-preferring p rats. Alcohol Clin Exp Res. 2010; 34(12): 2044–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrown AR, Herd MB, Belelli D, et al.: Developmentally regulated neurosteroid synthesis enhances GABAergic neurotransmission in mouse thalamocortical neurones. J Physiol. 2015; 593(1): 267–284. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCallachan H, Cottrell GA, Hather NY, et al.: Modulation of the GABAA receptor by progesterone metabolites. Proc R Soc Lond B Biol Sci. 1987; 231(1264): 359–69. PubMed Abstract | Publisher Full Text\n\nCarter RB, Wood PL, Wieland S, et al.: Characterization of the anticonvulsant properties of ganaxolone (CCD 1042; 3alpha-hydroxy-3beta-methyl-5alpha-pregnan-20-one), a selective, high-affinity, steroid modulator of the gamma-aminobutyric acid(A) receptor. J Pharmacol Exp Ther. 1997; 280(3): 1284–95. PubMed Abstract\n\nCefalu WT: Animal models of type 2 diabetes: clinical presentation and pathophysiological relevance to the human condition. ILAR J. 2006; 47(3): 186–98. PubMed Abstract | Publisher Full Text\n\nChen SR, Pan HL: Hypersensitivity of spinothalamic tract neurons associated with diabetic neuropathic pain in rats. J Neurophysiol. 2002; 87(6): 2726–33. PubMed Abstract\n\nChen H, Charlat O, Tartaglia LA, et al.: Evidence that the diabetes gene encodes the leptin receptor: identification of a mutation in the leptin receptor gene in db/db mice. Cell. 1996; 84(3): 491–5. PubMed Abstract | Publisher Full Text\n\nChisari M, Eisenman LN, Krishnan K, et al.: The influence of neuroactive steroid lipophilicity on GABAA receptor modulation: evidence for a low-affinity interaction. J Neurophysiol. 2009; 102(2): 1254–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChowdhury SK, Smith DR, Fernyhough P: The role of aberrant mitochondrial bioenergetics in diabetic neuropathy. Neurobiol Dis. 2013; 51: 56–65. PubMed Abstract | Publisher Full Text\n\nChung WK, Power-Kehoe L, Chua M, et al.: Genomic structure of the human OB receptor and identification of two novel intronic microsatellites. Genome Res. 1996; 6(12): 1192–9. PubMed Abstract | Publisher Full Text\n\nClascá F, Rubio-Garrido P, Jabaudon D: Unveiling the diversity of thalamocortical neuron subtypes. Eur J Neurosci. 2012; 35(10): 1524–32. PubMed Abstract | Publisher Full Text\n\nCooper A, Nutley M, MacLean EJ, et al.: Mutual induced fit in cyclodextrin-rocuronium complexes. Org Biomol Chem. 2005; 3(10): 1863–71. PubMed Abstract | Publisher Full Text\n\nCox CL, Huguenard JR, Prince DA: Nucleus reticularis neurons mediate diverse inhibitory effects in thalamus. Proc Natl Acad Sci U S A. 1997; 94(16): 8854–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nD'Amour FE, Smith DL: A method for determining loss of pain sensation. J Pharmacol Exp Ther. 1941; 72: 74–9. Reference Source\n\nD'Hulst C, Atack JR, Kooy RF: The complexity of the GABAA receptor shapes unique pharmacological profiles. Drug Discov Today. 2009; 14(17–18): 866–75. PubMed Abstract | Publisher Full Text\n\nDanaei G, Finucane MM, Lu Y, et al.: National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2·7 million participants. Lancet. 2011; 378(9785): 31–40. PubMed Abstract | Publisher Full Text\n\nDavis ME, Brewster ME: Cyclodextrin-based pharmaceutics: past, present and future. Nat Rev Drug Discov. 2004; 3(12): 1023–35. PubMed Abstract | Publisher Full Text\n\nDeFelipe J, Fariñas I: The pyramidal neuron of the cerebral cortex: morphological and chemical characteristics of the synaptic inputs. Prog Neurobiol. 1992; 39(6): 563–607. PubMed Abstract | Publisher Full Text\n\nDrel VR, Mashtalir N, Ilnytska O, et al.: The leptin-deficient (ob/ob) mouse: a new animal model of peripheral neuropathy of type 2 diabetes and obesity. Diabetes. 2006; 55(12): 3335–43. PubMed Abstract | Publisher Full Text\n\nEdwards JL, Vincent AM, Cheng HT, et al.: Diabetic neuropathy: mechanisms to management. Pharmacol Ther. 2008; 120(1): 1–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEvans MG, Marty A: Potentiation of muscarinic and alpha-adrenergic responses by an analogue of guanosine 5'-triphosphate. Proc Natl Acad Sci U S A. 1986; 83(11): 4099–103. PubMed Abstract | Free Full Text\n\nFernyhough P, Roy Chowdhury SK, Schmidt RE: Mitochondrial stress and the pathogenesis of diabetic neuropathy. Expert Rev Endocrinol Metab. 2010; 5(1): 39–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlor H, Bushnell MC: Central imaging of pain. In Hunt S, Koltzenburg M, eds. The neurobiology of pain. Oxford: Oxford University Press, 2005; 311–32. Publisher Full Text\n\nGentet LJ, Ulrich D: Strong, reliable and precise synaptic connections between thalamic relay cells and neurones of the nucleus reticularis in juvenile rats. J Physiol. 2003; 546(Pt 3): 801–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGredell JA, Turnquist PA, Maciver MB, et al.: Determination of diffusion and partition coefficients of propofol in rat brain tissue: implications for studies of drug action in vitro. Br J Anaesth. 2004; 93(6): 810–7. PubMed Abstract | Publisher Full Text\n\nGuillery RW, Harting JK: Structure and connections of the thalamic reticular nucleus: Advancing views over half a century. J Comp Neurol. 2003; 463(4): 360–71. PubMed Abstract | Publisher Full Text\n\nHosie AM, Wilkins ME, da Silva HM, et al.: Endogenous neurosteroids regulate GABAA receptors through two discrete transmembrane sites. Nature. 2006; 444(7118): 486–489. PubMed Abstract | Publisher Full Text\n\nHuh Y, Bhatt R, Jung D, et al.: Interactive responses of a thalamic neuron to formalin induced lasting pain in behaving mice. PLoS One. 2012; 7(1): e30699. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHumble SR: Dataset of ‘Neurosteroids are reduced in diabetic neuropathy and may be associated with the development of neuropathic pain’. Open Science Framework. 2016. Data Source\n\nJohnston GA: GABAA receptor channel pharmacology. Curr Pharm Des. 2005; 11(15): 1867–85. PubMed Abstract | Publisher Full Text\n\nJones BJ, Roberts DJ: The quantitative measurement of motor inco-ordination in naive mice using an accelerating rotarod. J Pharm Pharmacol. 1968; 20(4): 302–4. PubMed Abstract | Publisher Full Text\n\nKaplan JR, Wagner JD: Type 2 diabetes-an introduction to the development and use of animal models. ILAR J. 2006; 47(3): 181–5. PubMed Abstract | Publisher Full Text\n\nKeller AF, Coull JA, Chery N, et al.: Region-specific developmental specialization of GABA-glycine cosynapses in laminas I-II of the rat spinal dorsal horn. J Neurosci. 2001; 21(20): 7871–80. PubMed Abstract\n\nKeller AF, Breton JD, Schlichter R, et al.: Production of 5alpha-reduced neurosteroids is developmentally regulated and shapes GABAA miniature IPSCs in lamina II of the spinal cord. J Neurosci. 2004; 24(4): 907–15. PubMed Abstract | Publisher Full Text\n\nKnabl J, Witschi R, Hösl K, et al.: Reversal of pathological pain through specific spinal GABAA receptor subtypes. Nature. 2008; 451(7176): 330–4. PubMed Abstract | Publisher Full Text\n\nLambert JJ, Cooper MA, Simmons RD, et al.: Neurosteroids: endogenous allosteric modulators of GABAA receptors. Psychoneuroendocrinology. 2009; 34(Suppl 1): S48–S58. PubMed Abstract | Publisher Full Text\n\nLatham JR, Pathirathna S, Jagodic MM, et al.: Selective T-type calcium channel blockade alleviates hyperalgesia in ob/ob mice. Diabetes. 2009; 58(11): 2656–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi P, Shu HJ, Wang C, et al.: Neurosteroid migration to intracellular compartments reduces steroid concentration in the membrane and diminishes GABA-A receptor potentiation. J Physiol. 2007; 584(Pt 3): 789–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLindström P: The physiology of obese-hyperglycemic mice [ob/ob mice]. ScientificWorldJournal. 2007; 7: 666–85. PubMed Abstract | Publisher Full Text\n\nLübke J, Feldmeyer D: Excitatory signal flow and connectivity in a cortical column: focus on barrel cortex. Brain Struct Funct. 2007; 212(1): 3–17. PubMed Abstract | Publisher Full Text\n\nMaguire EP, Macpherson T, Swinny JD, et al.: Tonic inhibition of accumbal spiny neurons by extrasynaptic α4βδ GABAA receptors modulates the actions of psychostimulants. J Neurosci. 2014; 34(3): 823–838. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMayers JR, Iliff BW, Swoap SJ: Resveratrol treatment in mice does not elicit the bradycardia and hypothermia associated with calorie restriction. FASEB J. 2009; 23(4): 1032–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcFadyen MC, Melvin WT, Murray GI: Regional distribution of individual forms of cytochrome P450 mRNA in normal adult human brain. Biochem Pharmacol. 1998; 55(6): 825–30. PubMed Abstract | Publisher Full Text\n\nMensah-Nyagan AG, Meyer L, Schaeffer V, et al.: Evidence for a key role of steroids in the modulation of pain. Psychoneuroendocrinology. 2009; 34(Suppl 1): S169–S177. PubMed Abstract | Publisher Full Text\n\nMerskey H, Bogduk N: Classification of Chronic Pain. Descriptions of Chronic Pain Syndromes and Definitions of Pain Terms. Seattle: IASP Press, 1994. Reference Source\n\nMeyer HS, Schwarz D, Wimmer VC, et al.: Inhibitory interneurons in a cortical column form hot zones of inhibition in layers 2 and 5A. Proc Natl Acad Sci U S A. 2011; 108(40): 16807–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyer L, Patte-Mensah C, Taleb O, et al.: Allopregnanolone prevents and suppresses oxaliplatin-evoked painful neuropathy: multi-parametric assessment and direct evidence. Pain. 2011; 152(1): 170–81. PubMed Abstract | Publisher Full Text\n\nMeyer L, Patte-Mensah C, Taleb O, et al.: Cellular and functional evidence for a protective action of neurosteroids against vincristine chemotherapy-induced painful neuropathy. Cell Mol Life Sci. 2010; 67(17): 3017–34. PubMed Abstract | Publisher Full Text\n\nMiksys SL, Tyndale RF: Drug-metabolizing cytochrome P450s in the brain. J Psychiatry Neurosci. 2002; 27(6): 406–15. PubMed Abstract | Free Full Text\n\nMitchell EA, Gentet LJ, Dempster J, et al.: GABAA and glycine receptor-mediated transmission in rat lamina II neurones: relevance to the analgesic actions of neuroactive steroids. J Physiol. 2007; 583(Pt 3): 1021–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMogil JS: Animal models of pain: progress and challenges. Nat Rev Neurosci. 2009; 10(4): 283–94. PubMed Abstract | Publisher Full Text\n\nMountcastle VB: The columnar organization of the neocortex. Brain. 1997; 120(Pt 4): 701–22. PubMed Abstract | Publisher Full Text\n\nMunro G, Ahring PK, Mirza NR: Developing analgesics by enhancing spinal inhibition after injury: GABAA receptor subtypes as novel targets. Trends Pharmacol Sci. 2009; 30(9): 453–9. PubMed Abstract | Publisher Full Text\n\nPinault D, Deschênes M: Projection and innervation patterns of individual thalamic reticular axons in the thalamus of the adult rat: a three-dimensional, graphic, and morphometric analysis. J Comp Neurol. 1998; 391(2): 180–203. PubMed Abstract | Publisher Full Text\n\nPoisbeau P, Patte-Mensah C, Keller AF, et al.: Inflammatory pain upregulates spinal inhibition via endogenous neurosteroid production. J Neurosci. 2005; 25(50): 11768–76. PubMed Abstract | Publisher Full Text\n\nPritchett K, Mulder GB: The rotarod. Contemp Top Lab Anim Sci. 2003; 42(6): 49. PubMed Abstract\n\nRajalu M, Müller UC, Caley A, et al.: Plasticity of synaptic inhibition in mouse spinal cord lamina II neurons during early postnatal development and after inactivation of the glycine receptor alpha3 subunit gene. Eur J Neurosci. 2009; 30(12): 2284–92. PubMed Abstract | Publisher Full Text\n\nReddy DS, Rogawski MA: Ganaxolone suppression of behavioral and electrographic seizures in the mouse amygdala kindling model. Epilepsy Res. 2010; 89(2–3): 254–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanna E, Talani G, Busonero F, et al.: Brain steroidogenesis mediates ethanol modulation of GABAA receptor activity in rat hippocampus. J Neurosci. 2004; 24(29): 6521–6530. PubMed Abstract | Publisher Full Text\n\nSchlichter R, Keller AF, De Roo M, et al.: Fast nongenomic effects of steroids on synaptic transmission and role of endogenous neurosteroids in spinal pain pathways. J Mol Neurosci. 2006; 28(1): 33–51. PubMed Abstract | Publisher Full Text\n\nSchumacher M, Hussain R, Gago N, et al.: Progesterone synthesis in the nervous system: implications for myelination and myelin repair. Front Neurosci. 2012; 6: 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShu HJ, Eisenman LN, Jinadasa D, et al.: Slow actions of neuroactive steroids at GABAA receptors. J Neurosci. 2004; 24(30): 6667–75. PubMed Abstract | Publisher Full Text\n\nShu HJ, Zeng CM, Wang C, et al.: Cyclodextrins sequester neuroactive steroids and differentiate mechanisms that rate limit steroid actions. Br J Pharmacol. 2007; 150(2): 164–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSolovyova N, Moult PR, Milojkovic B, et al.: Bi-directional modulation of fast inhibitory synaptic transmission by leptin. J Neurochem. 2009; 108(1): 190–201. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStein DG: Progesterone exerts neuroprotective effects after brain injury. Brain Res Rev. 2008; 57(2): 386–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStoffel-Wagner B: Neurosteroid biosynthesis in the human brain and its clinical implications. Ann N Y Acad Sci. 2003; 1007: 64–78. PubMed Abstract | Publisher Full Text\n\nSullivan KA, Hayes JM, Wiggin TD, et al.: Mouse models of diabetic neuropathy. Neurobiol Dis. 2007; 28(3): 276–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTreede RD, Kenshalo DR, Gracely RH, et al.: The cortical representation of pain. Pain. 1999; 79(2–3): 105–11. PubMed Abstract | Publisher Full Text\n\nTsutsui K: Neurosteroids in the Purkinje cell: biosynthesis, mode of action and functional significance. Mol Neurobiol. 2008; 37(2–3): 116–25. PubMed Abstract | Publisher Full Text\n\nVareniuk I, Pavlov IA, Drel VR, et al.: Nitrosative stress and peripheral diabetic neuropathy in leptin-deficient (ob/ob) mice. Exp Neurol. 2007; 205(2): 425–36. PubMed Abstract | Publisher Full Text\n\nVincent AM, Edwards JL, McLean LL, et al.: Mitochondrial biogenesis and fission in axons in cell culture and animal models of diabetic neuropathy. Acta Neuropathol. 2010; 120(4): 477–89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvon Hehn CA, Baron R, Woolf CJ: Deconstructing the neuropathic pain phenotype to reveal neural mechanisms. Neuron. 2012; 73(4): 638–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZeilhofer HU: Loss of glycinergic and GABAergic inhibition in chronic pain--contributions of inflammation and microglia. Int Immunopharmacol. 2008; 8(2): 182–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15694",
"date": "15 Aug 2016",
"name": "Christopher Connolly",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI am not an electrophysiologist by training but am an expert in GABA(A) receptors.\nSummary:\nTIID mice have reduced neurosteroid tone but increased sensitivity to some neurosteroids. Ob/ob mice have mechanical hyperalgesia and allodynia (reduced by neurosteroid application).\nConclusions:\nReduced neurosteroid tone may be linked to hypersensitivity. Neurosteroids may exert analgesic effects by restoring GABAergic inhibitory tone.\n\nComments:\nI’m not clear on the comments relating to a developmental change in tone, yet the comment that this may be due to fluctuations in endogenous neurosteroid tone. I assume the authors do not mean to imply that development also fluctuates. Perhaps this needs rephrasing.\n\nIs anything known about the loss of neurosteroid tone? Is it due to altered neurosteroid production/degradation or receptor expression to non-responsive receptors?\n\nPresumably the use of the smaller cyclodextrins excludes the possibility of membrane cholesterol involvement. This should be stated.\n\nHas the time required for neurosteroid diffusion in a slice been confirmed using isolated neurons in culture (where no diffusion barrier exists)? This would help distinguish between neurosteroid-induced chronic changes in protein trafficking/gene expression and genuine diffusion if the changes can occur more quickly in culture.\n\nDoes the lack of finasteride on WT and ob/ob mice indicate that there is no endogenous basal neurosteroid-induced tone? The half-life of neurosteroid should be stated so that it may be considered with regard to being relevant.\n\nIt is interesting that galaxalone might be an allosteric modulator of GABA(A)Rs (as the other possibility of indomethacin as an antagonist has been excluded).\n\nThe rotarod test seems unfair when comparing fat mice to WT, but is valid in determining the detrimental effects of high doses of neurosteroid.\n\nThe effects of ganaxolone on vF is encouraging to indicate its potential relief from mechanical pain.\n\nThe possibility of differential binding sites to neurosteroids that could be differentially expressed is worthy of mention.\n\nWith the evidence presented here, I understand that the effects of Progesterone and DHP are greater in diabetic models, which would support an upregulation of the neurosteroid synthesis pathway as some compensatory mechanism. However, the evidence for the loss of neurosteroid tone remains elusive and probably upstream of the production of progesterone. This would presumably imply a deficit in mitochondrial cholesterol uptake, metabolism to pregnenolone, its efflux into the cytosol, or conversion to progesterone (from here on, the pathway to allopregnanolone is normal or enhanced). This needs to be spelled out more clearly.",
"responses": [
{
"c_id": "2500",
"date": "22 Feb 2017",
"name": "Stephen Humble",
"role": "Author Response",
"response": "Authors Comments in response to the reviewer Thank you very much indeed for talking the time to review the paper 1. The author agrees that development does not fluctuate. Neurosteroid tone starts out very high in the most immature neurons then decreases as with maturity. At around P20 the neurosteroid tone is almost negligible, but a modest neurosteroid tone has reappeared in mature neurones by P60 (Humble 2013, Brown 2012). The presence of the neurosteroid tone may be revealed with the use of cyclodextrin. Separately the overall τW of the mIPCs also decreases independently of the neurosteroid tone with progressive maturity (Humble 2013, Brown 2012). 2. Loss of neurosteroid tone in development thought to be due to reduced production. But separately there is also changes in the expression of different receptor subtypes (e.g. alpha-2 as per Bosman et al., 2005). 3. I can insert: ‘The use of small cyclodextrins was used to attempt to control for the generic effects of cyclodextrin such as an interaction with the cholesterol within the neuronal membrane.’ 4.The author had not performed experiments using isolated neurons in culture, nor explored specific potential changes in trafficking/gene expression. The lack of these experiments could be considered a relative limitation of the study and would be a useful complimentary study. The author used different methods of neurosteroid presentation in order to explore the issue of diffusion and ensure that neurosteroid penetration of the slice was optimal in each situation. In order to maximize the neurosteroid effect the slices were incubated more than an hour with the relevant neurosteroid compound as per work on other lipophilic GABAergic compounds by other authors (Benkwitz et al., 2007; Gredell et al., 2004) 5. The lack of effect of finasteride on WT and ob/ob mice is not sufficient to make the statement that there is no endogenous basal neurosteroid-induced tone. Indeed, there is in fact an endogenous basal neurosteroid-induced tone as illustrated by the experiments with cyclodextrin. The lack of observed effect with finasteride is consistent with other authors (Brown, 2012 Thesis). The lack of effect with finasteride in this context is explained as follows: Finasteride inhibits the production of new neurosteroid via 5α-reductase inhibition. However it does not impact on the endogenous neurosteroid compounds that are already present within the slice preparation. This question was considered during the project and it was decided that the appropriate way to answer this question would be to inject finasteride in vivo, some hours/days prior to the in vitro experiments. The half-life of the neurosteroid Allopregnanolone is approximately 30 minutes (Mellon et al., 2008). However, a functional equilibrium typically exists between Allopregnanlone and its neurosteroid precursors which act as an immediate reservoir that would not be blocked instantly by finasteride, which inhibits 5αR rather than 3α-HSD (Figure 1). 6. It is very interesting 7. Agreed. The rotarod test results for the ob/ob are included for completeness and for demonstrating the phenotype but have minimal validity other than that. The ob/ob rotarod experiments illustrate why the sedative effects of neurosteroid were studied in WT rather than the ob/ob. The ob/ob is type-2 diabetic because of obesity. Other obese mouse models such as the db/db also have type-2 diabetes. There does not appear to be an obese mouse model without type-2 diabetes. Thus the respective impacts of obesity and type-2 diabetes on rotarod ability cannot be differentiated. 8. This is a potentially translatable finding. 9. It is known that general affinity for the GABAA receptor varies between different neurosteroid compounds. Separately, GABAA receptors expressing specific subunits have greater affinity for a given neurosteroid. Indeed, the following is an extract from my thesis: ‘THIP is a low affinity partial agonist at receptors expressing the g subunit (i.e. most GABAARs), but has high affinity for GABAARs that express the a4b3d subunits where it behaves as a ‘superagonist’ compared to GABA (Brown et al., 2002; Farrant & Nusser, 2005; Krogsgaard-Larsen et al., 2004). d-GABAARs are predominantly located at extrasynaptic locations; therefore THIP is a useful pharmacological agent for the selective activation of tonic currents (Belelli et al., 2005; Farrant & Nusser, 2005).’ (Humble, 2013) For reasons of brevity the issue of differential binding sites to neurosteroids was not explored significantly within the paper, but this was covered in substantial detail within my thesis (Humble, 2013). It is summarised in the Conclusions section of the thesis: ‘Pipette-applied neurosteroids allopregnanolone and ganaxolone (but not DHP) were also able to prolong the decay time of cortical GABAAR mIPSCs in a differential manner. Allopregnanolone induced a greater effect than ganaxolone on GABAAR mIPSC decay time, but there was no difference in the response to the respective drugs between the WT and ob/ob mice. These findings suggest that neurosteroids are able to modulate the GABAAR from the intracellular compartment and that the sensitivity of the GABAAR is the same for WT and ob/ob mice. In addition, when indometacin was co-applied in the pipette with ganaxolone it had no impact on the efficacy of ganaxolone. This observation is inconsistent with the notion that indometacin competes with ganaxolone for its binding of the GABAAR.’ (Humble, 2013) 10. This is an excellent point and is the subject of a Brief Communication that I have written on the subject of Mitochondrial dysfunction and neurosteroid synthesis in diabetic neuropathy that will be published shortly. I would be happy to refer to this further in the discussion."
}
]
},
{
"id": "19599",
"date": "14 Feb 2017",
"name": "Anthony H. Dickenson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting study using electrophysiological and behavioural approaches in strains of mice to gauge the potential changes in neurosteroid signalling in diabetic neuropathy. The study extends through the central nervous system and the data is suggestive of important roles of these substances. The work is detailed and meticulous. However, using in vitro slices cannot allow for investigation of identified neurons in terms of pain inputs and how these may be altered by neurosteroids or indeed, neuropathy.",
"responses": [
{
"c_id": "2576",
"date": "23 Mar 2017",
"name": "Stephen Humble",
"role": "Author Response",
"response": "Dear Reviewer, thank you for all your comments!"
}
]
},
{
"id": "20664",
"date": "03 Mar 2017",
"name": "Xue-Jun Song",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the author investigated the neural electrophysiological characterizations of the spinal cord, the nucleus reticularis of the thalamus and the cerebral cortex by means of whole-cell patch-clamp recordings. The results suggested that diabetic mice showed reduced neurosteroid tone but enhanced sensitivity to some neurosteroids, ob/ob mice exhibited mechanical hyperalgesia and allodynia, which was reduced by neurosteroids applied exogenously. The author concluded that the reduced endogenous neurosteroid tone in ob/ob mice may be linked to their hypersensitivity.\n\nQuestions: 1. The author concluded the loss of neurosteroid tone in ob/ob mice. There is no further mechanism study. 2. What is the neurosteroid receptor expression in ob/ob mice, compared with WT mice? Is there any alternation in diabetic mice? 3.\n\nThe author did not show the frequency of mIPSC in lamina II of the spinal dorsal horn, nRT and layer 2/3 of the cortex. How about the frequency of mIPSC in those sites with development. Please show some representative consecutive chart recording. 4. According to the results, the amplitude of mIPSC is around 35-40 Pa, which is larger than our experience. Can the author give the reason. 5. Please can the author check the alphabetical sequences of the figures and figure legends.",
"responses": [
{
"c_id": "2577",
"date": "23 Mar 2017",
"name": "Stephen Humble",
"role": "Author Response",
"response": "1. Dear Reviewer, thank you for all your comments. This is an excellent point. The next logical step would be to carry out mass spectrometry to compare ob/ob and WT samples from pain pathway tissue. Unfortunately this was beyond the scope of the project. 2. GABA A receptors are the principle targets for the neurosteroid molecules that were investigated in these studies. The difference between ob/ob, db/db and WT mice is as follows: ob/ob lack leptin, db/db lack a functional leptin receptor and WT have normal leptin secretion and normal leptin receptor. The leptin receptor is entirely distinct from the GABA A receptor and there are no credible reports of leptin having a significant impact on the GABA A receptor. In anticipation of this question the second type 2 diabetic model (db/db) was used in order to illustrate that leptin was not implicated in the results. There is no specific reason a priori to consider that the GABA A receptor structure and function would be congenitally different than the WT. Indeed, the diabetic neuropathy happens with advancing maturity- hence why the experiments were carried out in mature animals that had developed neuropathic changes. The experiments using intracellular cyclodextrin illustrate that when the entire neurosteroid tone is removed by sequestration the baseline GABA A receptor function is the same for all strains of mice including diabetic and WT. Secondly, the application of the active neurosteroid allopregnanolone had the same impact on all strains. Thirdly, the intracellular application of neurosteroids enhanced GABA A receptor function by a similar margin in WT and ob/ob. All these observations taken together appear to be inconsistent with the hypothesis that GABA A receptor sensitivity differs between ob/ob, db/db and WT. If the GABA A receptor sensitivity is unchanged between strains this would be consistent with the hypothesis that GABA A receptor expression was not significantly different in the ob/ob, db/db and WT. 3. Due to the large amount of data contained within these studies the tables containing information such as the frequency of the mIPSCs are included as supplementary files. They can be accessed on the article’s webpage immediately below the Supplementary Material section and immediately above the References section. Any additional material not found there will be found in the author’s thesis using the link below: http://discovery.dundee.ac.uk/portal/en/theses/neurosteroids(c4659466-cd41-494d-aec6-edcf50e5274b).html In general terms, the frequency of mIPSCs increased with development in pain pathway neurons. This was anticipated due to known progressive synaptic development with maturation and was consistent with work from other authors. Specifically, there were no unexpected findings related to frequency with development. In addition, the neurosteroids investigated had no significant, consistent impact on frequency in the neurons studied. For these reasons, frequency has not been discussed in depth in the paper despite the fact that the frequency of the mIPSCs was recorded and analyzed routinely as part of the experiments along with other parameters. 4. This is an interesting observation. It is most likely due to differences in the utilization of the electrophysiological equipment and the process of analyzing the mIPSCs. It is not uncommon to observe differences between institutions for these reasons and there are many examples within the international literature. It is worth noting that the parameters such as amplitude, frequency, charge transfer, decay time in control recordings was consistent with recordings made by colleagues within the same institution that were using the same equipment set up and analysis protocols and software. In addition, it is standard established practice within the institution to exclude mIPSCs with a Rise Time greater than 1 millisecond (as mentioned in the Data Analysis section of the Methods). This approach excludes events (which have a longer Rise Time) because they are not immediately adjacent to the synapse. The author notes that the excluded events typically have lower amplitudes and have a morphology less like typical mIPSCs. 5. I have checked and rechecked the alphabetical sequences of the figures and figure legends and unfortunately cannot detect a specific issue. Please indicate which Figures and Figure Legends are the problem and I will correct them."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1923
|
https://f1000research.com/articles/5-194/v1
|
19 Feb 16
|
{
"type": "Research Article",
"title": "Surveillance, insecticide resistance and control of an invasive Aedes aegypti (Diptera: Culicidae) population in California",
"authors": [
"Anthony J. Cornel",
"Jodi Holeman",
"Catelyn C. Nieman",
"Yoosook Lee",
"Charles Smith",
"Mark Amorino",
"Katherine K. Brisco",
"Roberto Barrera",
"Gregory C. Lanzaro",
"F. Stephen Mulligan III",
"Anthony J. Cornel",
"Jodi Holeman",
"Catelyn C. Nieman",
"Charles Smith",
"Mark Amorino",
"Katherine K. Brisco",
"Roberto Barrera",
"Gregory C. Lanzaro",
"F. Stephen Mulligan III"
],
"abstract": "The invasion and subsequent establishment in California of Aedes aegypti in 2013 has created new challenges for local mosquito abatement and vector control districts. Studies were undertaken to identify effective and economical strategies to monitor the abundance and spread of this mosquito species as well as for its control. Overall, BG Sentinel (BGS) traps were found to be the most sensitive trap type to measure abundance and spread into new locations. Autocidal-Gravid-Ovitraps (AGO-B), when placed at a site for a week, performed equally to BGS in detecting the presence of female Ae. aegypti. Considering operational cost and our findings, we recommend use of BGS traps for surveillance in response to service requests especially in locations outside the known infestation area. We recommend AGO-Bs be placed at fixed sites, cleared and processed once a week to monitor mosquito abundance within a known infestation area. Long-term high density placements of AGO-Bs were found to show promise as an environmentally friendly trap-kill control strategy. California Ae. aegypti were found to be homozygous for the V1016I mutation in the voltage gated sodium channel gene, which is implicated to be involved in insecticide resistance. This strain originating from Clovis, California was resistant to some pyrethroids but not to deltamethrin in bottle bio-assays. Sentinel cage ultra-low-volume (ULV) trials using a new formulation of deltamethrin (DeltaGard®) demonstrated that it provided some control (average of 56% death in sentinel cages in a 91.4 m spray swath) after a single truck mounted aerial ULV application in residential areas.",
"keywords": [
"Aedes aegypti",
"surveillance",
"mosquito control",
"insecticide resistance",
"California",
"kdr"
],
"content": "Introduction\n\nAedes aegypti Linnaeus is presumed to have become established in the southeastern United States of America between the fifteenth and eighteenth centuries (Tabachnick, 1991). Its spread to the US between 1795 and 1905 precipitated major epidemics of yellow fever throughout the east coast and southern states (Crosby, 2006). Today this mosquito serves as the vector of three additional human viruses, dengue, chikungunya and Zika, which pose a major threat to global public health.\n\nThe state of California, however, had remained free from this vector until the summer of 2013 (Gloria-Soria et al., 2014), when Ae. aegypti were simultaneously collected in CO2-Baited Encephalitis Virus Surveillance traps (EVS) in the cities of Clovis (Fresno County) and Madera (Madera County) and later in oviposition cups in Menlo Park (San Mateo County). Within three months, Ae. aegypti, immature and adults were detected within a 1.6 km radius around the initial collection site in Clovis. An interactive guide tracking the 2013 and 2014 progression of the Ae. aegypti invasion in Clovis can be viewed as a webpage story developed by J. Holeman (http://bit.ly/1qB3CVD). Collections of Ae. aegypti in 2014 and 2015, and further expansion of its distribution proved that this mosquito species is capable of surviving through the winter and has established as a viable breeding population in California. By spring of 2015, this mosquito had been collected in seven additional California counties (Kern, Imperial, Los Angeles, Orange, Riverside, San Diego and Tulare). The ongoing widespread invasion and establishment of Ae. aegypti proves that this is a state-wide and not simply a regional issue in California.\n\nMultiple control measures were immediately implemented by the Consolidated Mosquito Abatement District (CMAD) in response to the initial discovery of Ae. aegypti in the city of Clovis. Control efforts included: (i) thorough property inspection for potential larval development sites, (ii) sanitation, (iii) insecticide applications to larval sources, (iv) residual barrier spraying with pyrethroid insecticides and (v) public education. Public education included distribution of information packets (consult www.mosquitobuzz.net for content) to all households within 512 meters of a positive collection site. Television and internet broadcasts and press stories were released throughout the period from initial detection to present day to generate public awareness.\n\nTo validate and improve Ae. aegypti surveillance and control, CMAD selected four traps from the wide variety available. These included: (i) the CO2- baited EVS trap (Rohe & Fall, 1979), (ii) oviposition cups (Barbosa et al., 2010; Furlow & Young, 1970), (iii) CO2- baited BG Sentinel (BGS) trap (www.bg-sentinel.com- without octenol attractant) and (iv) Autocidal Gravid Oviposition trap (AGO-B) (MacKay et al., 2013). The decision to use CO2 with the BGS was based on a study by de Ázara et al. (2013) that showed that more females were collected in BGS baited with CO2 than without.\n\nPyrethroid insecticides are the preferred method to control adult mosquitoes in California. However, there are numerous reports of pyrethroid resistant Ae. aegypti populations worldwide (Aponte et al., 2013; Hemingway & Ranson, 2000; Martins et al., 2009; Rawlins, 1998). Therefore the CMAD conducted insecticide susceptibility bio-assays on the Clovis population to determine their susceptibility to pyrethroids. In this study, bottle bio-assays were conducted exposing mosquitoes to pyrethrum, pyrethrum + piperonyl butoxide (PBO), permethrin, permethrin + PBO, sumithrin, deltamethrin and malathion. A portion of the voltage gated-sodium channel (vgsc) gene was sequenced in several mosquitoes to determine if well-known insecticide resistant mutations were present in the California Ae. aegypti. These genotypes confer resistance to both DDT and pyrethroid insecticides (Martins et al., 2009; Saavedra-Rodriguez et al., 2007). Sentinel cage mortality counts of Ae. aegypti were conducted in order to compare various ultra-low-volume (ULV) insecticide formulations. These involved the aerial delivery of insecticides from truck mounted sprays and were conducted in both open field and residential settings.\n\nIntegrated pest management (IPM) strategies that incorporate non-chemical based control methods are strongly recommended for mosquito control in California. Sustained high density placement of AGO-Bs, have shown promise as an effective Ae. aegypti trap-kill control measure in Puerto Rico (Barrera et al., 2014; MacKay et al., 2013). As part of fulfilling the IPM mission this non-chemical based strategy (using low density placement of AGO-Bs) was evaluated in Clovis.\n\n\nMethods\n\nA 10 × 10 cell grid covering 16.5 km2 (each cell represented a 0.16 km2) was selected for this study during summer 2013. The grid incorporated Ae. aegypti infested and non-infested areas (Figure 1). All four trap types were placed within the grid, which included 18 sites outside and 28 sites within the infestation area. The infestation area was defined as the area where Ae. aegypti had previously been recorded (within blue shaded area in Figure 1). The infestation area increased due to dispersal by the end of the study so that there were 34 positive trap sites in week 10.\n\nLocations of oviposition cups are represented by dark circles and AGO-Bs as stars. The blue shaded area corresponds to the area where Ae. aegypti were present at the start of the evaluation (infestation area). Traps are marked in red circles where Ae. aegypti were collected after commencement of the study, showing dispersal. Numbers correspond to trap site locations. Detailed trap count data for each of the 34 sites that collected mosquitoes at least once during the trial are provided in Supplementary figures S1–S7.\n\nOne AGO-B and one oviposition cup were placed approximately 50 m apart in each front yard site. These two traps were left continuously operational. Adult mosquitoes were counted and removed once a week from the AGO-B trap. Oviposition cups were checked weekly for the presence of eggs and a new sheet of oviposition paper (germination paper, Seedburo Equipment Company, Chicago, IL) was added. Each week the AGO-B and oviposition cup location were switched at each site. One night per week (1:00 pm to 8:00 am) an EVS and a BGS trap were placed 50 m apart in a property adjacent to each yard that had an AGO-B and oviposition cup. The EVS and BGS traps were rotated between each other every week. Mosquito numbers in the AGO-B traps were divided by seven to facilitate comparison of average trap night count with the EVS and BGS traps. Trap evaluations continued for nine weeks and both males and females were counted in the adult traps.\n\nSites selected within the infestation area were used to determine which of the three adult collecting trap types consistently collected the greatest number of adults (marked in stars in Figure 1). The goal was to identify the trap type most sensitive for ongoing Ae. aegypti surveillance. Adult mosquitoes were counted and removed from the AGO-B once every 7 days and from BGS and EVS 24 hours after each deployment. The sites outside the infestation area were used to determine which of the four trap types was most effective at first detecting Ae. aegypti dispersing out of the infestation area and therefore could be used to track dispersal of this mosquito.\n\nComparisons in numbers of Ae. aegypti adults collected by the different adult trap types were calculated for significant differences using the Wilcoxon-Rank-sum test (Bauer, 1972) implemented in the R statistical package version 3.0.0.\n\nLarvae of Ae. aegypti reared from eggs collected in oviposition cups in Clovis, California (CLOVIS strain) were used for bottle bio-assays and ULV trials. They were reared on a diet of ground rodent chow at 27°C under 14:10 hour (light:dark) photoperiod and adults were held at 70% relative humidity. The Rockefeller (ROCK) strain (Martins et al., 2009) was used as the susceptible Ae. aegypti strain and were reared under the same conditions. A pyrethroid-sensitive colony of Culex quinquefasciatus Say (CQ1), was used in one of the ULV trials as an additional pyrethroid susceptible control. The CQ1 strain was initially field-collected in Merced County, California, in the early 1950s and has been used at the Mosquito Control Research Laboratory, UC Davis as susceptible controls in insecticide bio-assays in the past (McAbee et al., 2004).\n\nTime to knockdown adulticide bottle bio-assays were conducted by treating the insides of 250 ml Wheaton bottles (Fisher #06-404B) with technical grade insecticides purchased from Chem Service (West Chester, PA). The insecticides were diluted in acetone and bottles were coated with the insecticide following the procedure described in Brogdon & McAllister (1998). For each insecticide, six replicates of 25 three to four day old adult mosquitoes were used to determine percentage mortality (malathion) and percent knock-down (pyrethroids) every 15 minutes for up to 2 hours and every 5 minutes between the 30th and 45th minutes. Control bottles were coated with acetone only. Mosquitoes that could not maintain an upright position when the bottle was rotated slowly were considered knocked down or dead. Mosquitoes were exposed to a predetermined dosage of insecticide that resulted in 100% mortality or knock-down within 30 minutes of the standard susceptible ROCK strain (Kuno, 2010). All bio-assays on the Clovis population were run simultaneously with the control susceptible ROCK strain. Concentrations of insecticide each test bottle was coated with were: Malathion = 50μg/ml; Deltamethrin = 10μg/ml; Sumithrin = 20μg/ml; Pyrethrum = 15.6μg/ml; Permethrin = 15μg/ml. The pyrethrum consisted of 14.2% pyrethrin I isomer and 10.7% pyrethrin II isomer (Lot # 2693200) and permethrin isomer ratio was 75.1% TRANS and 24.6% CIS (Lot # 3565000). The 400μg PBO per bottle dose used with pyrethrum and permethrin was the maximum amount that did not cause mortality when used alone. For bio-assays that included PBO, the mosquitoes were first exposed to PBO for one hour and then transferred to bottles coated with the insecticide.\n\nSignificance testing comparing 50% knock-down time (KD50) and 95% knock-down time (KD95) between the ROCK and CLOVIS strains exposed to the different chemicals and with and without PBO within strains were performed by Wilcoxon Rank-sum test (Bauer, 1972) using the R software package.\n\nThe IIS5-S6 region of the voltage gated sodium channel (vgsc) gene of 13 adult Ae. aegypti from Madera and 13 adults from Clovis, collected in BGS traps in the last week of August 2013, were sequenced using conventional Sanger Sequencing method. Samples were lysed using a Qiagen Tissulyser and genomic DNA extracted using a BioSprint 96 DNA Blood Kit (Qiagen, Chatsworth, CA) using the Qiagen BioSprint protocols described in Nieman et al. (2015). The PCR reaction was carried out following the protocol described in Martins et al. (2009). Amplicons were sequenced at the UC-DNA Sequencing Facility (College of Biological Sciences, UC Davis) using an ABI 3730 Genetic Analyzer (Applied Biosystems, Carlsbad, California). Gene fragments were also sequenced in both directions (forward/reverse) and SNPs were identified only if the SNP was found in both directions. Geneious (Kearse et al., 2012) software version 6.1.4 was used for sequence alignment and SNP identification.\n\nAdult mosquitoes were exposed under operational field conditions to the following commercial ULV adulticide formulations; 6% pyrethrins, 60% piperonyl butoxide (PBO) (Evergreen EC® 60-6, MGK, Minneapolis, MN); etofenprox (Zenivex® E20, Wellmark International, Schaumburg, IL); and deltamethrin (DeltaGard®, Bayer, Research Triangle Park, NC). Evergreen EC® 60-6 and Zenivex® E20 are registered in California. DeltaGard® is not presently registered in California and a Research Authorization (approved RA-1505051) was obtained from the CA Department of Pesticide Regulation for evaluating this product for this study. Application rates and relevant meteorological conditions during applications of the three ULV trials are provided in Table 1. All three ULV applications were evaluated in a fallow open field and the trial using deltamethrin was also performed in a residential area within the city of Clovis. Mosquito control efficacy results were based on 12 hour post exposure mortalities recorded in sentinel cages placed in rows perpendicular to the wind direction and downwind from the line of application. Approximately 20 CLOVIS, ROCK and CQ1 mosquitoes were placed in screened sentinel cages (Townzen & Natvig, 1973) 3 to 6 hours prior to the ULV trial. Mosquitoes in the sentinel cages were provided access to a cotton swab soaked with a 10% sucrose solution and held in a cool environment in insulated boxes for transport to the field. Within 30 minutes prior to the commencement of the trial, sentinel cages were attached to stakes 1 m above ground. The stakes were placed in the ground 15.25, 30.48, 60.96 and 91.44 m downwind from an application in the open setting. Stakes holding sentinel cages in the residential setting were positioned in the configuration depicted in Figure 2. This configuration was designed to assess penetration of the ULV (91.4 m swath) in the urban residential. All applications were made with a truck mounted, cold aerosol ULV sprayer (Cougar model with SmartFlow, Clarke, Roselle, ILL). Controls were placed in an area away from the spray sites. Sentinel cages were left on the stakes for an hour post application, after which knock-down and mortality was recorded in each cage. The mosquitoes were left in the cages, and each cage was covered on one side with a lightly dampened towel, the cotton swabs were re-soaked with 10% sucrose and each cage was individually placed into a plastic bag and held for a further 12 hours in insulated boxes. After 12 hours, mortality was recorded in each cage. Sentinel cages from the control sites were treated in exactly the same manner.\n\nDistance between the sentinel cages from the street to the furthest sentinel in the front yard of the next parallel street was 91.44 m.\n\nZenivex® was applied at 4g/ha, MGK Pyrocide® at 60g/ha and DeltaGard® at 1.5g/ha. A temperature inversion of 1°C and wind speeds of 2.41 to 4.83 km/hr were recorded during the open field applications. A temperature inversion of 0.3°C and wind speeds 8.85 to 12.55 km/hr were recorded during the trial in the Clovis residential area. CQ1 stands for Cx. quinquefasciatus Johannesburg strain.\n\na Zenivex®,;\n\nb MGK Pyrocide®,\n\nc and d DeltaGard®\n\nTwo glass microscope slides (Bioquip, Rancho Dominguez, CA) mounted on spinners (Hock Company, Gainesville, FL) adjacent to the sentinel cages were used to record droplet size and density of passing airborne spray across 91.44 m in both the open and residential settings. Teflon coated slides were used for all ULV trials except for the trials using deltamethrin which were coated with magnesium oxide. DropVision® was used as the software system to read the slides and generate the droplet analysis reports (Leading Edge Associates, LLC out of Waynesville, NC). Slides were digitally read using a specialized Motic DMBA300 Teflon slide reading compound microscope (Leading Edge Associates Inc.) at 100X magnification.\n\nRegression line slope calculations were performed to examine if there was any difference in mortalities of mosquitoes placed at various distances from the spray sources to test for distance effect. Wilcoxon rank-sum tests were performed to compare mortalities of mosquitoes in sentinel cages across the 91.44 m swath for each mosquito strain exposed to each ULV formulation. Both the regression line slopes (Chambers, 1992) and Wilcoxon rank-sum tests (Bauer, 1972) were calculated using the corresponding option in the R software package.\n\nAs a trap-kill system, the AGO-B was designed to capture female Ae. aegypti on a sticky surface as they entered the trap to oviposit (Barrera et al., 2014). To evaluate this trap-kill control concept, three general locations within the Clovis Ae. aegypti infestation area were selected (Figure 3). Each of the three locations were more than 200 meters apart, a distance further than the typical distance Ae. aegypti fly (60–100 m, Harrington et al., 2005; Valerio et al., 2012). Within the intervention area (area A in Figure 3), one AGO-B was placed in the front yard of each of 144 households. In this study, one AGO-B was placed at each parcel in contrast to three AGO-Bs per parcel in Barrera et al. (2014). Six BGS were deployed within the two control areas (areas B and C in Figure 3) and monitored for a 12 week period; two weeks before and four weeks after deployment of the AGO-Bs within area A (treatment site). Three BGS were also used to monitor mosquito numbers within the treatment site for 12 weeks. Trap counts from an additional fifteen AGO-Bs positioned outside the treatment area (dark circles in Figure 3) were also included in the study to represent control area female AGO-B counts. Female mosquitoes were counted in AGO-Bs once a week in the treatment and control areas, and male and female Ae. aegypti were counted in BGS twice a week for the duration of the trial.\n\nProperties shaded in grey were those that had an AGO-B trap placed in their front yard (144 traps in total). Stars represent locations of BGS traps used to measure abundance per week in both the control and intervention areas. Dark circles show positions of AGO-B traps outside the intervention area which were used for control site monitoring.\n\nRelative temporal abundance comparisons of Ae. aegypti in the AGO-Bs and BGS traps between the treatments (area A) and control sites (areas B and C), were calculated as normalized proportions per week per area, by dividing the number of mosquitoes trapped per week by the total number of mosquitoes collected from the corresponding trap over the six weeks. Normalizing done to BGS trap counts in the control areas B and C were combined. Regression line slopes and statistical significance were calculated using a linear model function, lm (Wilkinson & Rogers, 1973), in the R statistics package.\n\n\nResults\n\nNumbers of males and females collected in the various trap types within the infestation area varied considerably from week to week (Dataset 1). Mean numbers and SD of Ae. aegypti collected at each site in each trap is given above each bar in Figures S1 to S7 (Dataset 1). Overall the BGS traps collected the largest number of adult Ae. aegypti and these were collected at significantly (P<0.005) more sites than either the AGO-B or EVS traps. However, the AGO-B and BGS traps performed equally, with no significant difference (P>0.05), in detection of female Ae. aegypti (Figure 4, Wilcoxon rank sum test P-value=0.23). BGS traps collected more males than AGO-B traps in most weeks (Figures S2 and S4, Dataset 1).\n\nData included the 34 traps numbered in Figure 1.\n\nOf the 18 sites outside the infestation area, only six recorded the presence and therefore spread of Ae. aegypti during the 10 week trial period. In three of these sites both AGO-B and BGS traps collected mosquitoes in the same week. In the other three sites only the BGS collected Ae. aegypti. None of the EVS and oviposition cup traps outside the infestation area collected Ae. aegypti adults or eggs. The sites that captured mosquitoes outside the original infestation area are represented by the red circles in Figure 1.\n\nThe DNA fragment containing the IIS5-S6 region of vgsc had identical nucleotide sequences among all samples from Madera and Clovis (GenBank accession: KU728155-6). See Dataset 1 for sequence and alignment to the Liverpool strain reference sequence. All Madera and Clovis Ae. aegypti were homozygous for the known pyrethroid resistant V1016I mutation. The intron between exons 20 and 21 of California Ae. aegypti were 15 bp longer than the reference genome with 73% sequence similarity to the reference strain. In addition to the sequence difference in V1016I (exon 21), there were two other synonymous nucleotide differences in exons 20 between the California mosquitoes and the reference genomes (amino acid position 981 and 982).\n\nIn the standard bottle bio-assays no mortality was observed with either strain in the control bottles. Also no mortality was observed in mosquitoes due to PBO exposure for one hour before placement into bottles coated with the pyrethroids. The KD50 and KD95 times for the CLOVIS strain were highly variable between the six bottle replicates coated with sumithrin and pyrethrum and to a lesser extent with permethrin (Figure 5) but were still significantly longer than the ROCK strain (P=0.0032). Exposing the CLOVIS mosquitoes to PBO for one hour significantly narrowed and slightly shortened their mean KD50 and KD90 to times closer to that experienced by the ROCK strain (Figure 5). Both the CLOVIS and ROCK strains produced similar knock-down times and mortality against deltamethrin and malathion respectively (Figure 5).\n\nROCK refers to the susceptible Rockefeller colony strain and CLOVIS refers to mosquitoes found in Clovis. Concentrations of insecticide each test bottle was coated with were: Malathion = 50μg/ml; Deltamethrin = 10μg/ml; Sumithrin = 20μg/ml; Pyrethrum = 15.6μg/ml; Permethrin = 15μg/ml. Significant differences in knock-down time or mortality between the CLOVIS and ROCK strains is indicated by red text (Wilcoxon rank sum test α<0.05 or P<0.0032). * Significant differences in knock-down time between treatment with and without PBO (Wilcoxon rank sum test α<0.05 or P < 0.0032).\n\nLess than 10% mortality occurred in sentinel cages positioned in the up-wind control site locations for all trials. Mortality in sentinels exposed to insecticides was corrected for natural mortality by Abbott’s formula (Abbott, 1925). No insecticide droplets were observed on the slides at the control sites in the open and residential ULV trials. The average droplet size and distribution recorded at all distances in ULV trials fell within the recommended range of 5–25μm for ground ULV applications (Bonds, 2012).\n\nThere was no significant decline in mortality over distance from spray source (linear model P>0.05; Table 1) in sentinel caged mosquitoes in any of the open field ULV trials. Because there was no significant mortality effect in distance from spray source within the swath (91.44 m) we combined mortalities in all sentinels to produce a single average mortality across the swath. Applications of pyrethrum + PBO, etofenprox and deltamethrin in open settings resulted in 100% mortality at all distances from the spray source up to 91.44 m in both the susceptible ROCK (Ae. aegypti) and CQ1 (Cx. quinquefasciatus) mosquito strains (Table 1). The mortality rate in the CLOVIS strain was significantly lower than the ROCK strain for both etofenprox (Wilcoxon rank sum test P=0.00021) and pyrethrum + PBO (P=0.00073) applications (Table 1). In the open field ULV deltamethrin application 100% mortality was achieved with all three strains, including the CLOVIS strain, confirming the bottle bio-assay data that Clovis Ae. aegypti were susceptible to deltamethrin. In the ULV trial conducted in the residential area, there was also no significant decline in mortality over distance from spray source in sentinel caged mosquitoes (Table 1). However, the CLOVIS strain had significantly lower mortality than the ROCK strain (Wilcoxon rank-sum test P=6.26×10-5). At the dosage of deltamethrin applied, which was less than half of the maximum allowable, according to the label, we achieved almost 99.1% mortality of CQ1 mosquitoes at all four distances along the swath (Table 1). Mortality of the Clovis Ae. aegypti was considerably less with an average rate of 55.64% along the full swath length.\n\nCounts of Ae. aegypti males and females collected in BGS and AGO-B traps in the treatment and control sites are provided in Figures S8–S10 (Dataset 1). During the 12 week study period in 2014, the six BGS traps in the control areas collected 650 males (mean= 5.75 [SD=6.54] per trap night) and 1035 females (mean= 9.16 [SD=6.55] per trap night). The 15 control AGO-B traps collected 1,189 females (mean = 6.6 [SD=6.2] per week). During this six week period when the AGO-B traps were deployed, there was no significant decline in Ae. aegypti collected in the BGS traps (slope= -0.0006; P=0.643) or AGO-B traps (slope=0.0020; P=0.536) within the control areas (Figure 6A and 6C). However, during the same period, a decline in Ae. aegypti counts in BGS traps in the intervention area was significant (slope= -0.0047; P= 0.036; Figure 6B) and a decline in Ae. aegypti in the AGO-B traps in the intervention area was significant (slope= -0.0035; P=0.002; Figure 6D).\n\nRelative abundance of females and males were calculated in BGS traps and only females in AGO-B traps.\n\n\nDiscussion\n\nDuring the dry summer breeding period (May–October) in Clovis, the major source of water to sustain breeding of Ae. aegypti is water accumulation in small containers and refuse from residential watering. In other dry urban locations where Ae. aegypti is found, such as in Arizona, watering by homeowners and monsoonal summer rainfall create sources of water for breeding. Based on the oviposition cup data within the infestation area during the 10 week trap evaluation trial the average numbers of eggs deposited in oviposition cups in Clovis per week was 291 (SD= 432.6), which was less than the average of 447.6 eggs/day in Tucson, Arizona (Hoeck et al., 2003). The average number of female Ae. aegypti in BGS traps per night were 4 individuals in the primary Clovis infestation area and a similar abundance was observed between the summer months (June–September) of 2013 and 2014. Average BGS trap counts in Clovis were similar to the average of 4.67 per BGS trap night counts in Cairns, Australia (Williams et al., 2006) which were both lower than the average per night trap count of 58.8 females in BGS traps in Florida (Wright et al., 2015). Average numbers of female Ae. aegypti per week in AGO-B traps in Puerto Rico of 3.83 (Barrera et al., 2014) is higher than the average numbers collected in Clovis which was 2 mosquitoes per week in the 34 traps deployed during the 10 week trap evaluation trial in 2013 and about 2.75 mosquitoes per week in the 15 AGO-Bs deployed in the control areas during the 12 week AGO-B control evaluation trial in 2014.\n\nDespite the uniform residential setting in Clovis, variable temporal abundance of adult Ae. aegypti was observed in this study regardless of trap type (Figure S1, Dataset 1). The highly variable spatial and temporal numbers of mosquitoes collected in BGS traps in Clovis is typical for Ae. aegypti trapping dynamics in general (Degener et al., 2014; Williams et al., 2007). This clustering and variation in numbers needs to be taken into account by public health and mosquito control agencies when monitoring abundance over time, even in relatively small areas. In this study, the spatial design of trap placements was not appropriate to measure clustering and aggregate effects of Ae. aegypti in Clovis. However, clustering of Ae. aegypti typically does occur in residential areas (Williams et al., 2006) and this also needs to be considered in the design of monitoring strategies. Williams et al. (2007) recommended use of square-root transformations rather than log data transformations to deal with non-normally distributed BGS trap count data.\n\nIn this study, BGS traps out-performed the other three trap types in measuring both the spread and abundance of Ae. aegypti in Clovis. However, purchase and operational costs of BGS traps and homeowner cooperation in placement of traps must be considered in trap selection and use. Consequently, the CMAD now utilizes the general surveillance strategy described as follows: BGS traps are deployed in response to public service requests, particularly in locations outside known infestation areas to get a quick but sensitive measure of Ae. aegypti presence and to document dispersal. To correct for daily fluctuations in trap collections that could cause a missed detection of Ae. aegypti, AGO-B traps are also deployed at some properties outside the known infestation areas. AGO-B traps are deployed for a week or longer as opposed to BGS traps which are set out for only one day. Some AGO-B traps are also placed at fixed sites to monitor general abundance within the known infestation area. As an augmentative measure, oviposition cups, which are less expensive and less time consuming to utilize, are used at fixed points both within and outside the infestation area. Deployment of oviposition cups is often less conspicuous and may generate less homeowner concern and greater acceptance.\n\nThe bottle bio-assay data provided clear evidence that the Clovis Ae. aegypti population is resistant to some pyrethroids such as permethrin, sumithrin and pyrethrum. All the Clovis and Madera Ae. aegypti sequenced, were fixed for the V1016I amino acid substitution which is one of the knock-down resistance mutations in vgsc responsible for reduced sensitivity to pyrethroids (Saavedra-Rodriguez et al., 2007). Multiple amino acid substitutions, associated with pyrethroid resistance, clustered within the II24-S5 linker, 11S5-S6 helices and the corresponding regions of domain III of the sodium channel gene (Vontas et al., 2012) have been found in various populations of Ae. aegypti worldwide, and they also include other mutations such as V1016G (Brengues et al., 2003; Chang et al., 2009) and F1534C (Harris et al., 2010). Addition of PBO in bottle bio-assays reversed resistance to pyrethrum significantly (P<0.003 ; Figure 5), suggesting that some P450s were additionally responsible for conferring metabolic resistance to pyrethrum in Clovis Ae. aegypti. The addition of PBO to permethrin did not significantly reduce knock-down time (Figure 5). The wide range of knock-down times in response to permethrin and pyrethrum exposures indicate that the detoxifying role of P450s was variable between individuals and hence is likely a genetically polymorphic trait among Clovis Ae. aegypti.\n\nInterestingly, despite being fixed for the V1016I mutation and having indications of the presence of the P450 metabolic pathway, the Clovis Ae. aegypti were not resistant to the pyrethroid deltamethrin. Presence of the V1016G and F1534C substitutions and other metabolic mechanisms associated with pyrethroid resistance in Ae. aegypti (Vontas et al., 2012) have yet to be found in the Ae. aegypti introduced into California.\n\nIncreased susceptibility to pyrethrum by addition of PBO warranted evaluating the control efficacy of a synergized pyrethrum + PBO formulation in a field ULV trial situation. In the open line application with no obstruction to the material drift, all the ROCK strain died but only 57.9% of CLOVIS were killed in sentinel cages within a 91.44 m swath (Table 2). The low mortality of Clovis Ae. aegypti in the ULV trial was unexpected because there was a strong synergizing effect observed in the bottle bio-assays (Figure 5). Pyrthrum + PBO formulations are favored for ULV control in California because of labeling which allows application over agricultural crops. Unfortunately, results from this study indicate that pyrethrum + PBO ULV formulations may not control Ae. aegypti in Clovis.\n\nVertebrate toxicity effects of PBO are of concern to the public and two other ULV pyrethroid formulations with no PBO were evaluated in field ULV trials. Higher mortality was achieved with etofenprox in the Clovis Ae. aegypti in an open ULV trial (75%). Bottle bio-assays were not performed with etofenprox because specific crystallization properties of this chemical prevent it from coating surfaces evenly, and an even coating of the bottles is required for consistency of bottle bio-assay results. High mortality of Clovis Ae. aegypti (100%) was achieved with deltamethrin (DeltaGard®) in the open ULV trial, as was expected due to supportive low knock-down times observed in the bottle bio-assays (Figure 5). Application of DeltaGard® in a residential Clovis setting resulted in lower but still promising 57.3% mortality of sentinel caged Clovis Ae. aegypti after exposure. Based on these results, we believe that use of this formulation may be effective to achieve significant immediate suppression of adult females in disease epidemic situations when applied in multiple consecutive nights as recommended by Macedo et al. (2010). Differences in mortality in Clovis Ae. aegypti between open and residential applications were likely due to reduced spread and penetration of the aerosolized product around residential structures and landscapes and less optimal local meteorological conditions. The most preferable time for ULV applications in the southern San Joaquin Valley of California is generally at sunset, when temperature inversions and wind conditions are most favorable for achieving the required 91.44 m swath insecticide drift. However, this is a peak time of day for human activity in residential areas. Timing of application may prove somewhat of a limiting factor for routine use of ground based adulticide application efforts against Ae. aegypti.\n\nWe observed a gradual decline in Ae. aegypti counts in areas where a single AGO-B trap was deployed at every household as an intervention. These results suggest that long-term, high density placement of AGO-B traps could be effective in Clovis. We speculate that deployment of multiple (3-4) AGO-B traps per parcel, similar to Barrera et al. (2014), might reduce Ae. aegypti populations below nuisance or disease transmission levels.\n\n\nConclusion\n\nWe provided much needed information regarding the effective and economical strategies of surveillance and control for the Zika and other arbovirus vector, Ae. aegypti. Considering operational cost and our control research results, we recommend use of BGS traps for surveillance for Ae. aegypti in locations where presence of Ae. aegypti has not been recorded. AGO-Bs can be used as a surveillance tool within a known infestation area. Long-term high density placements of AGO-Bs were found to show promise as an environmentally friendly trap-kill control strategy. We recommend conducting insecticide resistance assays of Aedes aegypti populations wherever they exist because their susceptibility to insecticides differ geographically. Our surveillance and control methods can be applied to other closely related species such as Aedes albopictus which also transmits arboviruses and share similar biology. Given that Ae. aegypti transmits multiple serious viral diseases to humans, it is strongly recommended to include mosquito control research to monitor and develop effective control strategies.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data (Figures S1–S11) for 'Surveillance, insecticide resistance and control of an invasive Aedes aegypti (Diptera: Culicidae) population in California', 10.5256/f1000research.8107.d114301 (Cornel et al., 2016).",
"appendix": "Author contributions\n\n\n\nAJC, SM and JH conceived the study, designed all the experiments and conducted field work. KB, MA and CS assisted in field work. CN extracted DNA and sequenced a portion of the voltage-gated-sodium channel gene. YL performed data analysis. YL, AJC and JH and KB made figures for the manuscript. RB donated the AGO-B traps and assisted in design of the experiments using these traps. AJC prepared the first draft, and all authors were involved in the revision of the draft manuscript and have agreed to final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe thank the Fresno Mosquito and Vector Control District for providing some funds to conduct the insecticide resistance assays. DeltaGard® product used in the Ultra-Low-Volume trials was donated by Bayer (Research Triangle Park, NC) and we thank Dennis Candito (ADAPCO Inc- Sanford, FL) for calibrating the truck mounted ULV applicators and Chris Olsen (Bayer) and Gary Braness (Yosemite Environmental Services, Fresno, CA) for setting up spinners and measuring droplet sizes in the DeltaGard® ULV trials. We thank CDC at Puerto-Rico for donating the AGO-B traps. We are grateful to Valkyrie Kimball (Marin Sonoma Mosquito and Vector Control District) for measuring droplet sizes and densities for the etofenprox and pyrethrin ULV trials. We thank the board of trustees of the Consolidated Mosquito Abatement District for approving funds allocated to the rest of the study.\n\n\nReferences\n\nAbbott WS: A method of computing the effectiveness of an insecticide. J Econ Entomol. 1925; 18(2): 265–267. Publisher Full Text\n\nAponte HA, Penilla RP, Dzul-Manzanilla F, et al.: The pyrethroid resistance status and mechanisms in Aedes aegypti from the Guerrero state, Mexico. Pestic Biochem Physiol. 2013; 107(2): 226–234. Publisher Full Text\n\nBarbosa AAC, Favaro EA, Mondini A, et al.: Evaluation of oviposition traps as an entomological surveillance method for Aedes aegypti (Diptera: Culicidae). Rev Bras Entomol. 2010; 54(2): 328–331. Publisher Full Text\n\nBarrera R, Amador M, Acevedo V, et al.: Sustained, area-wide control of Aedes aegypti using CDC autocidal gravid ovitraps. Am J Trop Med Hyg. 2014; 91(6): 1269–1276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBauer DF: Constructing confidence sets using rank statistics. J Am Statistical Assoc. 1972; 67(339): 687–690. Publisher Full Text\n\nBonds JA: Ultra-low-volume space sprays in mosquito control: a critical review. Med Vet Entomol. 2012; 26(2): 121–130. PubMed Abstract | Publisher Full Text\n\nBrengues C, Hawkes NJ, Chandre F, et al.: Pyrethroid and DDT cross-resistance in Aedes aegypti is correlated with novel mutations in the voltage-gated sodium channel gene. Med Vet Entomol. 2003; 17(1): 87–94. PubMed Abstract | Publisher Full Text\n\nBrogdon WG, McAllister JC: Simplification of adult mosquito bioassays through use of time-mortality determinations in glass bottles. J Am Mosq Control Assoc. 1998; 14(2): 159–164. PubMed Abstract\n\nChambers JM: Linear models. Chapter 4 of Statistical Models in S. eds., JM Chambers and TJ Hastie, Wadsworth & Brooks/Cole, 1992.\n\nChang C, Shen WK, Wang TT, et al.: A novel amino acid substitution in a voltage-gated sodium channel is associated with knockdown resistance to permethrin in Aedes aegypti. Insect Biochem Mol Biol. 2009; 39(4): 272–278. PubMed Abstract | Publisher Full Text\n\nCornel AJ, Holeman J, Nieman CC, et al.: Dataset 1 in: Surveillance, insecticide resistance and control of an invasive Aedes aegypti (Diptera: Culicidae) population in California. F1000Research. 2016. Data Source\n\nCrosby MC: The American plague: the untold story of yellow fever, the epidemic that shaped our history. 1st ed. Berkley Books, New York, 2006. Reference Source\n\nDegener CM, Eiras AE, Azara TM, et al.: Evaluation of the effectiveness of mass trapping with BG-Sentinel traps for Dengue vector control: a cluster randomized controlled trial in Manaus, Brazil. J Med Entomol. 2014; 51(2): 408–420. PubMed Abstract | Publisher Full Text\n\nde Ázara TM, Degener CM, Roque RA, et al.: The impact of CO2 on collection of Aedes aegypti (Linnaeus) and Culex quinquefasciatus Say by BG-Sentinel® traps in Manaus, Brazil. Mem Inst Oswaldo Cruz. Rio de Janeiro, 2013; 108(2): 229–232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFurlow BM, Young WW: Larval surveys compared to ovitrap surveys for detecting Aedes aegypti and Aedes triseriatus. Mosq News. 1970; 30(3): 468–470. Reference Source\n\nGloria-Soria A, Brown JE, Kramer V, et al.: Origin of the dengue fever mosquito, Aedes aegypti, in California. PLoS Negl Trop Dis. 2014; 8(7): e3029. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarrington LC, Scott TW, Lerdthusnee K, et al.: Dispersal of the dengue vector Aedes aegypti within and between rural communities. Am J Trop Med Hyg. 2005; 72(2): 209–220. PubMed Abstract\n\nHarris AF, Rajatileka S, Ranson H: Pyrethroid resistance in Aedes aegypti from Grand Cayman. Am J Trop Med Hyg. 2010; 83(2): 277–284. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHemingway J, Ranson H: Insecticide resistance in insect vectors of human disease. Annu Rev Entomol. 2000; 45: 371–391. PubMed Abstract | Publisher Full Text\n\nHoeck PA, Ramburg FB, Merrill SA, et al.: Population and parity levels of Aedes aegypti collected in Tucson. J Vector Ecol. 2003; 28(1): 1–8. PubMed Abstract\n\nKearse M, Moir R, Wilson A, et al.: Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012; 28(12): 1647–1649. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuno G: Early history of laboratory breeding of Aedes aegypti (Diptera: Culicidae) focusing on the origins and use of selected strains. J Med Entomol. 2010; 47(6): 957–971. PubMed Abstract | Publisher Full Text\n\nMacedo PA, Schleier JJ 3rd, Reed M, et al.: Evaluation of efficacy and human health risk of aerial Ultra-low volume applications of pyrethrins and piperonyl butoxide for adult mosquito management in response to West Nile virus activity in Sacramento County, California. J Am Mosq Control Assoc. 2010; 26(1): 57–66. PubMed Abstract | Publisher Full Text\n\nMackay AJ, Amador M, Barrera R: An improved autocidal gravid ovitrap for the control and surveillance of Aedes aegypti. Parasit Vectors. 2013; 6(1): 225. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartins AJ, Lins RM, Linns JG, et al.: Voltage-Gated Sodium Channel Polymorphism and Metabolic Resistance in Pyrethroid-Resistant Aedes aegypti from Brazil. Am J Trop Med Hyg. 2009; 81(1): 108–115. PubMed Abstract\n\nMcAbee RD, Kang KD, Stanich MA, et al.: Pyrethroid tolerance in Culex pipiens pipiens var molestus from Marin County, California. Pest Manag Sci. 2004; 60(4): 359–368. PubMed Abstract | Publisher Full Text\n\nNieman CC, Yamasaki Y, Collier TC, et al.: A DNA extraction protocol for improved DNA yield from individual mosquitoes [version 1; referees: 3 approved]. F1000Res. 2015; 4: 1314. Publisher Full Text\n\nRawlins SC: Spatial distribution of insecticide resistance in Caribbean populations of Aedes aegypti and its significance. Rev Panam Salud Publica. 1998; 4(4): 243–251. PubMed Abstract | Publisher Full Text\n\nRohe DL, Fall RP: A miniature battery powered C02 baited light trap for mosquito borne encephalitis. Bull Soc Vector Ecol. 1979; 4: 24–27.\n\nSaavedra-Rodriguez K, Urdaneta-Marquez L, Rajatileka S, et al.: A mutation in the voltage-gated sodium channel gene associated with pyrethroid resistance in Latin American Aedes aegypti. Insect Mol Biol. 2007; 16(6): 785–798. PubMed Abstract | Publisher Full Text\n\nTabachnick WJ: Evolutionary genetics and insect borne disease. The yellow fever mosquito, Aedes aegypti. Am Entomol. 1991; 37: 14–24.\n\nTownzen KR, Natvig HL: A disposable adult mosquito bioassay cage. Mosq News. 1973; 33(1): 113–114. Reference Source\n\nValerio L, Facchinelli L, Ramsey JM, et al.: Dispersal of male Aedes aegypti in a coastal village in southern Mexico. Am J Trop Med Hyg. 2012; 86(4): 665–676. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVontas J, Kioulos E, Pavlidi N, et al.: Insecticide resistance in the major dengue vectors Aedes albopictus and Aedes aegypti. Pest Biochem Physiol. 2012; 104(2): 126–131. Publisher Full Text\n\nWilkinson GN, Rogers CE: Symbolic descriptions of factorial models for analysis of variance. J R Stat Soc Ser C Appl Stat. 1973; 22(3): 392–9. Publisher Full Text\n\nWilliams CR, Long SA, Russell RC, et al.: Field efficacy of the BG-Sentinel compared with CDC Backpack Aspirators and CO2-baited EVS traps for collection of adult Aedes aegypti in Cairns, Queensland, Australia. J Am Mosq Control Assoc. 2006; 22(2): 296–300. PubMed Abstract | Publisher Full Text\n\nWilliams CR, Long SA, Webb CE, et al.: Aedes aegypti population sampling using BG-Sentinel traps in north Queensland Australia: statistical considerations for trap deployment and sampling strategy. J Med Entomol. 2007; 44(2): 345–350. PubMed Abstract | Publisher Full Text\n\nWright JA, Larson RT, Richardson AG, et al.: Comparison of BG-Sentinel® Trap and Oviposition Cups for Aedes aegypti and Aedes albopictus Surveillance in Jacksonville, Florida, USA. J Am Mosq Control Assoc. 2015; 31(1): 26–31. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "12576",
"date": "01 Mar 2016",
"name": "Saul Lozano-Fuentes",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI read this research with great enjoyment. The authors take a well-rounded approach to the surveillance and control of Aedes aegypti. However, I have a few methodological observations.Methodologically, the calculation of the Knockdown Time (KDT) is commonly done using a binary logistic regression. From the text is unclear but I assumed that the authors are using the 50th and 95th percentiles for the estimation of the KDT50 and KDT95. Box plots are non-parametric; they are used without making any assumptions of underlying statistical distributions. The authors should make use of existing commonly used models to describe their time response observations. Several journals including, the Journal of Medical Entomology, request the use of logistic regression with a logit or probit link function. Using a common methodology allows the comparison of time response bioassay from different places and years. Thusly, please also provide estimates and confidence intervals for the KDT50 and KDT90 values using a binary logistic regression. Values beyond the KDT90 are outside the linear portion of the regression. This publication could help you quickly calculate these statistics Lozano-Fuentes et al1.; other options exist in prepackage solutions like the “DRC” r library (https://cran.r-project.org/web/packages/drc/drc.pdf)Regarding the V1016I mutation, the authors provide evidence that the mutation is fixed in their samples and it is very likely in at a high frequency in the local population. Please add Vera-Maloof et al.2 to your discussion to help frame this discovery with closer Ae. aegypti populations.Regarding the AGO-Bs trap evaluation, the provided evidence (Figure 6, panel D) does not show a significant difference in mean relative abundance between any of the weeks as shown by the overlapping confidence intervals, some of the intervals are large enough that they appear to be below zero. Then again, the authors present a significant negative slope. These conflicting results point to the variables not being normal or not having equal variances. Please provide evidence that the data does not break the underlying assumptions of the linear regression model since it is being used to support the claim of abundance reduction by the AGO-B.Other obsevationsPlease add the citation to the R software to your manuscript 3.",
"responses": [
{
"c_id": "1841",
"date": "02 Mar 2016",
"name": "Yoosook Lee",
"role": "Reader Comment",
"response": "Thank you very much for the constructive criticism. We'll make our revision soon to accommodate your suggestions. I would like to clarify one point, though, that our estimate of KD50 and KD95 was from binary logistic regression implemented in MASS library. We acknowledge that this is omitted in our v1 manuscript. The estimates were provided in Supplemental Table 1. Because there are too many lines overlapping and direct clear comparion was difficult using traditional time response plot. Therefore we decided to plot the distribution of KD50 and KD95 (will revise to KD90 in version 2) to better visualize the comparison between chemicals. We will revise to use KD90 instead of KD95 and clarify our method of calculating KD50 and KD90 in the version 2."
}
]
}
] | 1
|
https://f1000research.com/articles/5-194
|
https://f1000research.com/articles/5-1921/v1
|
05 Aug 16
|
{
"type": "Research Article",
"title": "Synthesis, characterization and toxicity studies of pyridinecarboxaldehydes and L-tryptophan derived Schiff bases and corresponding copper (II) complexes",
"authors": [
"Margarita Malakyan",
"Nelly Babayan",
"Ruzanna Grigoryan",
"Natalya Sarkisyan",
"Vahan Tonoyan",
"Davit Tadevosyan",
"Vladimir Matosyan",
"Rouben Aroutiounian",
"Arsen Arakelyan",
"Nelly Babayan",
"Ruzanna Grigoryan",
"Natalya Sarkisyan",
"Vahan Tonoyan",
"Davit Tadevosyan",
"Vladimir Matosyan",
"Rouben Aroutiounian"
],
"abstract": "Schiff bases and their metal-complexes are versatile compounds exhibiting a broad range of biological activities and thus actively used in the drug development process. The aim of the present study was the synthesis and characterization of new Schiff bases and their copper (II) complexes, derived from L-tryptophan and isomeric (2-; 3-; 4-) pyridinecarboxaldehydes, as well as the assessment of their toxicity in vitro. The optimal conditions of the Schiff base synthesis resulting in up to 75-85% yield of target products were identified. The structure-activity relationship analysis indicated that the location of the carboxaldehyde group at 2-, 3- or 4-position with regard to nitrogen of the pyridine ring in aldehyde component of the L-tryptophan derivative Schiff bases and corresponding copper complexes essentially change the biological activity of the compounds. The carboxaldehyde group at 2- and 4-positions leads to the higher cytotoxic activity, than that of at 3-position, and the presence of the copper in the complexes increases the cytotoxicity. Based on toxicity classification data, the compounds with non-toxic profile were identified, which can be used as new entities in the drug development process using Schiff base scaffold.",
"keywords": [
"Schiff base",
"L-tryptophan",
"copper (II) complex",
"synthesis",
"cytotoxicity",
"HeLa",
"KCL-22"
],
"content": "Introduction\n\nSchiff bases are considered as a very important class of organic ligands having a wide range of applications in many fields of biomedicine1–3. They are the condensation products of an amino compound with an active carbonyl compound and carry imine or azomethine (–C=N–) functional group, which is essential for their biological activity. Structurally, Schiff base is a nitrogen analog of an aldehyde or ketone, in which the carbonyl group (C=O) has been replaced by an imine or azomethine group. While aliphatic aldehyde containing Schiff bases are unstable in nature and readily get polymerized, aromatic aldehyde containing compounds are more stable due to conjugation system4.\n\nSchiff bases derived from aromatic aldehydes and aromatic amines are widely applicable in the fields of biology, inorganic and analytical chemistry5,6. Their biological activities are based on the earlier detected anti-inflammatory, antiviral, antibacterial, antifungal, antimalarial and antipyretic properties7–10. The bonding interactions between aromatic amino acid side chains of the receptor and aromatic/heteroaromatic rings of the ligand were revealed in most of X-ray crystal structures of protein complexes with small molecules. This protein-ligand recognition, based on aromatic ring involved non-covalent interactions, can ensure the application of Schiff bases, derived from aromatic aldehydes and amines, in drug design process11–13. Moreover, the evaluation of the structure-activity relationship of Schiff bases, derived from different substituted aromatic amines and aldehydes, demonstrated the importance of the latter for desired biological activity14–15.\n\nPyridinecarboxaldehyde derivatives of Schiff bases are of great interest because of their role in natural and synthetic organic chemistry. It is known, that pyridoxal-amino acid systems are important in numerous metabolic reactions intermediated with amino acid and pyridoxal. So far, pyridinecarboxaldehyde isomers characterized by different localization of carboxaldehyde group (2-, 3- or 4-) relative to nitrogen atom in pyridine ring are valuable precursors for complex forming Schiff bases, since they can exhibit physiological effects similar to pyridoxal-amino acid systems. Thus, the pyridinecarboxaldehydes containing Schiff bases are expected to have enhanced biological activities.\n\nIt is known that the binding of bioorganic molecules or drugs to metal ions drastically change their biomimetic properties, therapeutic effects, and pharmacological properties16. Schiff base derivatives of aromatic amino acids are good chelating agents and capable to form stable complexes with transition metals and exhibit significant biological and enzymatic activities17,18. The most widely studied cation in this respect is copper, which is implicated in a wide range of vital cell functions. Copper has been proven to be beneficial against several diseases such as tuberculosis, rheumatoid, gastric ulcers and cancers19–21. Many non-toxic, lipid-soluble, small molecular mass copper chelate complexes have been shown to have superoxide dismutase- and catalase-like activities22–24, which makes them essential for de novo syntheses of metalloelement-dependent enzymes required for oxygen utilization and prevention of oxygen superoxide accumulation.\n\nThe present study describes the synthesis and characterization of isomeric 2-, 3- and 4-pyridinecarboxaldehydes and L-tryptophan derived Schiff bases and their copper (II) complexes, as well as the assessment of their cytotoxic activity.\n\n\nMethods\n\nAll chemicals and solvents used were of analytical grade. The reagents used for the synthesis of the Schiff bases and their copper (II) complexes were obtained from Sigma-Aldrich (Sigma-Aldrich Co. LLC, USA), including L-tryptophan, 2-; 3-; and 4-pyridinecarboxaldehydes, KOH, copper (II) acetate, and methanol.\n\nSchiff bases (2pyr.Trp, 3pyr.Trp, and 4pyr.Trp) were synthesized by condensation of potassium salt of L-tryptophan and isomeric (2-, 3-, 4-) pyridinecarboxaldehydes, respectively, in alcohol solutions (ethanol, methanol) in 5°C–25°C temperature range and 1:1 molar ratio. First, 10 mM of L-tryptophan was dissolved in 100 mL of alcohol solution, containing KOH (10mM) by permanent stirring under dry nitrogen at 18°C–20°C. Then 10 mM of the corresponding isomer of pyridincarboxyaldehyde (2-, 3- or 4-) was added to the resulting solution with stirring and refluxed at 50°C for 2 hours resulting in a yellow colored solution that indicates the Schiff base formation. The volume of the solution was then reduced in vacuo using a rotary evaporator. Anhydrous ether was added to deposit a yellowish precipitate, which was then re-crystallized from alcohol.\n\nThe obtained Schiff bases were served as ligands for the synthesis of appropriate copper (II) complexes (namely, Cu-2pyr.Trp, Cu-3pyr.Trp, and Cu-4pyr.Trp). The synthesis was performed at 20±2°С in alcohol media (methanol, ethanol) using potassium hydroxide and copper acetate. Complex formation was carried out in a reaction medium without preliminary isolation of Schiff bases. Compound isolation was performed by partial evaporation of the solvent, settling, centrifugation, re-crystallization, and vacuum drying.\n\nFor characterization of Schiff bases and their copper (II) complexes the infrared (IR) absorbance spectra were obtained in the range of 4000–400 cm-1s in Vaseline oil on KBr plates using Spectrometer IR 75 (Carl Zeiss, Jena). For assessment of thermal stability of obtained compounds the IR absorbance spectra were recorded every 30 minutes in the 60°С – 100оС temperature range for 2 hours.\n\nThe elemental analysis was performed by combustion in a pure oxygen environment using PerkinElmer 2400 Series II CHNS/O Elemental Analyzer (PerkinElmer, USA). The nuclear magnetic resonance (NMR) spectra were obtained in D2O and CD3OD on Spectrometer Varian 300 MHz (Agilent, USA).\n\nSolubility assessment of synthesized compounds was carried out according to standard test method protocol25. Schiff bases were water-soluble, while their copper (II) complexes were soluble in dimethyl sulfoxide (DMSO). Stock solutions of Schiff bases and their copper (II) complexes at the concentration of 10 mM/mL were prepared and diluted with nutrition medium RPMI-1640 or DMEM. Only freshly prepared solutions were used in experiments.\n\nHuman HeLa (cervix carcinoma) and KCL-22 (chronic myeloid leukemia) cell lines were obtained from the cell culture collection of the Institute of Molecular Biology (Yerevan, Armenia). Growth media (DMEM and RPMI-1640) as well as media supplements were obtained from Sigma-Aldrich. The human HeLa and KCL-22 cell lines were routinely maintained at 37°C in the growth medium DMEM (HeLa) and RPMI-1640 (KCL-22), supplemented with 10% fetal bovine serum (HyClone, UK), 2 mM L-glutamine (Sigma Aldrich, Germany), 100 IU/mL penicillin (Sigma Aldrich, Germany) and 100 μg/mL streptomycin (Sigma Aldrich, Germany).\n\nThe cytotoxicity of test compounds was assessed using standard protocols for Trypan blue exclusion test and neutral red uptake (NRU) assay26,27.\n\nTrypan blue exclusion test: The KCL-22 cells were seeded into 15 ml glass vials at the density of 0.5 × 106 cells/mL. After 48 hours the test compounds were added at the concentrations of 0.1 µM/mL, 1 µM/mL, 10 µM/mL, 100 µM/mL, and 1000 µM/mL. After further incubation for 48 hours, cells were stained with 0.4% Trypan blue solution for 5-15 minutes and counted in a haemocytometer under a light microscope. The viable cell number was determined.\n\nNRU assay: The HeLa cells were seeded at the density of 0.3 × 106 cells/mL into 96-well plates (Corning, USA), incubated for 48 hours, and then test compounds were added to the cell cultures at the concentrations of 0.1 µM/mL, 1 µM/mL, 10 µM/mL, 100 µM/mL, and 1000 µM/mL. After further incubation for 48 hours the NRU assay was performed. The absorbance was measured using a microplate reader (Human Reader HS, Germany) at a wavelength of 570 nm.\n\nCell viability was expressed as a percentage of the negative control (cell cultures with no treatment). Doses inducing 50% inhibition of cell viability (the IC50 value) were calculated to determine the cytotoxicity of Schiff bases and their copper(II) complexes.\n\nThe extrapolation of obtained IC50 values into LD50 values for compound acute rodent oral toxicity in vivo was performed. The regression formula was used to weigh up the starting doses for single oral application in rats28:\n\nBased on obtained LD50 values the class of toxicity was assigned to all synthesized compounds29. Since the human cell lines were used for the experiments, the IC50 values obtained have also prognostic significance for human30.\n\nAll experiments were done in at least three replicates. At least triplicate cultures were scored for an experimental point. All values were expressed as means ± SE. The Student’s one tailed t-test was applied for statistical analysis of results, p < 0.05 was considered as the statistically significant.\n\n\nResults and Discussion\n\nThe optimal conditions of the Schiff base synthesis with the use of potassium salt of L-tryptophan were identified, allowing to obtain the yield of target products up to 85% for 2pyr.Trp, 75% for 3pyr.Trp, and 80% for 4pyr.Trp. Based on 1H NMR spectra inspection for synthesized Schiff bases in D2O and CD3OD, the loss of 2 singlet signals of СН2 group was revealed, confirming the presence of target compounds. The observed 1Н NMR spectral data, particularly, the presence of CH signal (duplet-duplets) of –СН2–СН– in the range of 4.1–4.5 ppm are characteristic for Schiff bases. The results of elemental analysis of 2pyr.Trp, 3pyr.Trp and 4pyr.Trp Schiff bases are presented in the Table 1.\n\nThe brutto chemical formula for 2pyr.Trp, 3pyr.Trpand 4pyr.Trp Schiff bases was identified as C17H14N3O2K (Mr=331.41). The suggested structure of synthesized Schiff bases is presented in the Figure 1. Decomposition temperature of 2pyr.Trp, 3pyr.Trp, and 4pyr.Trp Schiff bases was in the range of 180ºC – 190ºC.\n\nSuggested structure of the 2pyr.Trp (A), 3pyr.Trp (B) and 4pyr.Trp (C) Schiff bases.\n\nObtained Schiff bases were then used for the synthesis of corresponding copper (II) complexes (Cu-2pyr.Trp, Cu-3pyr.Trp, Cu-4pyr.Trp). Based on IR analysis the shift of IR absorbance bands upon formation of copper metallocomplexes with Schiff bases was observed (Table 2). Upon formation of Schiff bases the valence deviation (NH) of the tryptophan indole ring was shifted from 3430cm-1 to 3180-3190cm-1 due to the intramolecular interaction of indole and pyridine rings. In copper metallocomplexes this band appeared in the area of 3250–3270cm-1, which was associated with the changes in conjugation linkage degree (C=N) with pyridine ring caused by coordination bond Cu....N. This in turn resulted in changes of interactions between the pyridine and indole rings. Coordination bond Cu....N caused also a shift of valence deviations (C=N) towards low frequencies, and this band was practically overlapped by the band of valence deviations of (C=О-). The results of the determination of the copper content in metallocomplexes by atomic absorption and elemental analysis of carbon, nitrogen and hydrogen are presented in the Table 3. The obtained data allowed to suggest that metallocomplexes contain two Schiff base ligands and have brutto formula C34H28N6O4Cu (Mr = 648.17). The inferred structures of Cu-2pyr.Trp, Cu-3pyr.Trp, Cu-4pyr.Trp metallocomplexes are presented in the Figure 2. Thermostability assessment demonstrated no changes in IR spectra at 100оС during 2 hours. Decomposition temperature for these compounds was at a range of 180оC – 190оC without melting.\n\nSuggested structure of Cu-2pyr.Trp (A), Cu-3pyr.Trp (B), Cu-4pyr.Trp (C) metallocomplexes.\n\nThe synthesized Schiff bases and their copper complexes were tested in vitro to determine their cytotoxicity in Hela and KCL-22 cell lines. Our results indicate that the cytotoxic activity of 2pyr.Trp and its copper (II) complex depends on a cell line (Figure 3, A and B, Dataset 1, File 1). In case of 2pyr.Trp action in HeLa cell line, a hormesis effect was apparent, since a significant increase in the cell number was observed at lower concentrations of 0.1-1µM/mL. Further dosage increase, however, did not lead to the total cell death, since the cell viability was more than 90% compared to untreated cells, even at the highest concentration tested (1000 µM/mL). In contrast, the KCL-22 cell line was more sensitive against cytotoxicity of 2pyr.Trp; the hormesis effect was only slightly visible, but the further dose-dependent cell viability decrease was observed and resulting in 20% of cell viability at the highest tested concentration (1000 µM/mL) (Figure 3, A). In case of Cu-2pyr.Trp the sensitivity of cell lines was reversed. Starting from 10 µM/mL concentration the viability of HeLa cells decreased substantially compared with the KCL-22 cell line (Figure 3, B).\n\nThe cytotoxicity of 2pyr.Trp (A) and Cu-2pyr.Trp (B) in HeLa and KCL-22 cell lines. Dose-response curves were obtained after 48 hours of treatment with Schiff base 2pyr.Trp and its copper(II) complex Cu-2pyr.Trp at the concentration range of 0.1–1000 µM/mL. Cell viability was expressed as a percentage of the negative control (cell cultures with no treatment). Doses inducing 50% inhibition of cell viability (the IC50 value) were calculated to determine the cytotoxicity of 2pyr.Trp and Cu-2pyr.Trp. The IC50 value estimated for 2pyr.Trp in KCL-22 cell line was equal to 56±9.1 μM/mL, whereas the viability of HeLa cells was more than 90% at the highest concentration tested (A). The IC50 values estimated for Cu-2pyr.Trp were equal to 7±1.7 μM/mL and 80±7.5 μM/mL for HeLa and KCL-22 cell lines, respectively (B).\n\nThe non-cytotoxic profile was observed for Schiff base 3pyr.Trp in both cell lines (Figure 4, A, Dataset 1, File 2), while, Cu-3pyr.Trp demonstrated the increased cytotoxic activity against both cell lines. (Figure 4, B). However, the IC50 value was possible to estimate only for HeLa cells, since the viability of KCL-22 cells was more than 60% at the highest concentration tested (1000 µM/mL).\n\nThe cytotoxicity of 3pyr.Trp (A) and Cu-3pyr.Trp (B) in HeLa and KCL-22 cell lines. Dose-response curves were obtained after 48 hours of treatment with Schiff base 3pyr.Trp and its copper(II) complex Cu-3pyr.Trp at the concentration range of 0.1–1000 µM/mL. Cell viability was expressed as a percentage of the negative control (cell cultures with no treatment). Doses inducing 50% inhibition of cell viability (the IC50 value) were calculated to determine the cytotoxicity of 3pyr.Trp and Cu-3pyr.Trp. The non-cytotoxic profile was observed for 3pyr.Trp in both cell lines, since the viability of HeLa and KCL-22 cells was around 100% at the highest concentration tested (A). The Cu-3pyr.Trp demonstrated the increased cytotoxic activity against both cell lines, however, the IC50 value was possible to estimate only for HeLa cells (500±5.6 μM/mL), since the viability of KCL-22 cells was more than 60% at the highest concentration tested (B).\n\nThe toxicity profile of 4pyr.Trp (Figure 5, A and B, Dataset 1, F3) was similar to 2pyr.Trp, since the slight hormesis effect was again evident in HeLa cell line, while the KCL-22 cells were more sensitive against its cytotoxic activity (Figure 5, A). Despite these similarities, the overall cytotoxic activity of 4pyr.Trp was higher compared to 2pyr.Trp. The level of cell viability at the highest tested concentration (1000 µM/mL) for 4pyr.Trp (Figure 5, A) were 70% for HeLa and 12% for KCL-22, respectively, while in case of 2pyr.Trp viability levels were 90% (HeLa) and 30% (KCL-22), respectively (Figure 4, A). Again, HeLa cells were more susceptible to the cytotoxicity of Cu-4pyr.Trp than KCL-22 cells (Figure 5, B). Earlier, several Schiff bases were tested in vitro for their cytotoxic activity against different cell lines and the structure activity relationship of compounds was discussed17,31,32. Furthermore, Kril et al. reported on the hormesis effect for MCF-7 and 647-V tumour cells, and suggested that receptor-mediated mechanisms are responsible for the observed phenomenon. Here, we also demonstrated the hormesis effect in HeLa cell line, which supports the statement about potential receptor-binding ability of Schiff bases due to the presence of carbon–nitrogen double bond. The changes in cytotoxic activity against cancer cell lines were shown earlier depending on the presence of different groups (chloro, methoxy, nitro, and phenyl) in aromatic rings of a Schiff base molecule32. Furthermore, we have noted that even the localization of carboxaldehyde group at 2-, 3- or 4-position with regard to nitrogen of aromatic ring can affect the cytotoxicity of Schiff bases.\n\nThe cytotoxicity of 4pyr.Trp (A) and Cu-4pyr.Trp (B) in HeLa and KCL-22 cell lines. Dose-response curves were obtained after 48 hours of treatment with Schiff base 4pyr.Trp and its copper(II) complex Cu-4pyr.Trp at the concentration range of 0.1–1000 µM/mL. Cell viability was expressed as a percentage of the negative control (cell cultures with no treatment). Doses inducing 50% inhibition of cell viability (the IC50 value) were calculated to determine the cytotoxicity of 4pyr.Trp and Cu-4pyr.Trp. The IC50 value estimated for 4pyr.Trp in KCL-22 cell line was equal to 100±6.5 μM/mL, whereas the viability of HeLa cells was more than 60% at the highest concentration tested (A). The IC50 values estimated for Cu-4pyr.Trp were equal to 10±5 μM/mL and 30±3.8 μM/mL for HeLa and KCL-22 cell lines, respectively (B).\n\nBased on dose-response curves the half-maximal inhibitory concentrations (IC50 values) were estimated for Schiff bases and their copper (II) complexes (Table 4). Our data suggest that Schiff bases 2pyr.Trp, 3pyr.Trp and 4pyr.Trp are non-toxic for HeLa cells since the IC50 values were impossible to estimate even at the highest tested concentration (1000 µM/mL). The 2pyr.Trp (IC50=56±9.1 μM/mL) was two times more toxic against the KCL-22 cell line, than 4pyr.Trp (IC50=100±6.5 μM/mL), while the 3pyr.Trp demonstrated the same non-toxic profile as it was shown in HeLa cells.\n\nThe cytotoxic activity was observed for Cu-2pyr.Trp, Cu-3pyr.Trp and Cu-4pyr.Trp in HeLa cell line with the IC50 values of 7±1.7 μM/mL, 500±5.6 μM/mL and 10±5.0 μM/mL, respectively. Those IC50 values were significantly lower in comparison with their copper free analogs. The same tendency was demonstrated for Cu-4pyr.Trp in KCL-22 cell line, where the IC50 value decreased up to 30±3.7 μM/mL. In case of Cu-2pyr.Trp and Cu-3pyr.Trp complexes tested in KCL-22 cell line, no significant differences in IC50 values were observed in comparison with their copper free analogs. Thus, it can be assumed that the cytotoxic activity of Schiff bases tends to increase at complex formation with the copper molecule (Figure 6).\n\n*p<0.01.\n\nTesting of the compounds’ effects on the viability of cells grown in culture is widely used as a predictor of potential toxic effects in whole animals28. Our extrapolated data on the predicted LD50 doses demonstrated that the tested compounds 3pyr.Trp, 4pyr.Trp, and Cu-3pyr.Trp belong to the Class IV of non-toxic chemicals, while 2pyr.Trp, Cu-2pyr.Trp, and Cu-4pyr.Trp belong to the Class III of slightly toxic compounds (Table 4)29. Since the human cell lines were used for the experiments, those hazard classification data have also a prognostic significance for the human. The United States Food and Drug Administration (FDA) states that it is essential to perform toxicological studies during the development of new drugs, since the desirable pharmacological activity needs to be achieved in the absence of acute toxicity33. The non-toxic profile of 3pyr.Trp, 4pyr.Trp and Cu-3pyr.Trp Schiff bases indicates that this compounds can be considered as new entities in drug development process.\n\n\nConclusions\n\nWe have synthesized and characterized several new Schiff bases of aromatic amino acid derivatives and their copper complexes. Cytotoxicity tests indicated that 3pyr.Trp, 4pyr.Trp, and Cu-3pyr.Trp are non-toxic for human, whereas compounds 2pyr.Trp, Cu-2pyr.Trp, and Cu-4pyr.Trp retain slight toxicity. Moreover, obtained results indicate that cell lines HeLa (epithelial origin) and KCL-22 (derived from blood) vary in sensitivity to the cytotoxic action of the tested compounds; the latter suggests the tissue-/cell line-specificity of their effect. The results also demonstrate that structural alterations, namely, the localization of the carboxaldehyde group at 2-, 3- or 4-position with regard to nitrogen of pyridine ring in aldehyde component of the L-tryptophan derivative Schiff bases and corresponding copper complexes essentially change the biological activity of the compounds tested. The carboxaldehyde group at 2- and 4-positions leads to the higher cytotoxic activity, than that of at 3-position, the presence of the copper in the complexes, mostly increases the cytotoxicity. Thus, the results obtained may be used for the further development of pharmaceutical agents based on L-tryptophan and pyridinecarboxaldehyde derived Schiff bases and their copper(II) complexes.\n\n\nData availability\n\nF1000Research: Dataset 1. raw data of generated dose-response curves, 10.5256/f1000research.9226.d13023534",
"appendix": "Author contributions\n\n\n\nMM, RA and AA formulated the initial project and aims. VT, DT and VM carried out the synthesis and characterization of the tested compounds. NB contributed to the experimental design of toxicity studies, RG and NS carried out the toxicity studies. MM, NB and AA prepared the first draft of the manuscript. All authors contributed to the review of the manuscript and agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAuthors acknowledge funding from the International Science and Technology Center (ISTC) in the frames of the projects A-1764 (MM) and A-2116 (MM and AA).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Dr. Gennadi Gasparyan for his advice and expertise that greatly assisted the research.\n\n\nReferences\n\nZoubi WAl: Biological Activities of Schiff Bases and Their Complexes: A Review of Recent Works. Int J Org Chem. 2013; 3: 73–95. Publisher Full Text\n\nArulmurugan S, Kavitha HP, Venkatraman BR: Biological activities of Schiff base and its complexes: A review. Rasayan J Chem. 2010; 3(3): 385–410. Reference Source\n\nKajal A, Bala S, Kamboj S, et al.: Schiff Bases: A Versatile Pharmacophore. J Catal. 2013; 2013: 1–14. Publisher Full Text\n\nPaul P: Ruthenium, osmium and rhodium complexes of polypyridyl ligands: Metal-promoted activities, stereochemical aspects and electrochemical properties. J Chem Sci. 2002; 114(4): 269–276. Publisher Full Text\n\nCimerman Z, Miljanić S, Galić N: Schiff Bases Derived from Aminopyridines as Spectrofluorimetric Analytical Reagents. Croat Chem Acta. 2000; 73(1): 81–95. Reference Source\n\nKabak M, Elmali A, Elerman Y, et al.: Conformational study and structure of bis-N,N′-p-bromo-salicylideneamine-1,2-diaminobenzene. J Mol Struct. 2000; 553(1–3): 187–192. Publisher Full Text\n\nPrzybylski P, Huczynski A, Pyta K, et al.: Biological Properties of Schiff Bases and Azo Derivatives of Phenols. Curr Org Chem. 2009; 13(2): 124–148. Publisher Full Text\n\nda Silva CM, da Silva DL, Modolo LV, et al.: Schiff bases: A short review of their antimicrobial activities. J Adv Res. 2011; 2(1): 1–8. Publisher Full Text\n\nSaid SA, Amr Ael-G, Sabry NM, et al.: Analgesic, anticonvulsant and anti-inflammatory activities of some synthesized benzodiazipine, triazolopyrimidine and bis-imide derivatives. Eur J Med Chem. 2009; 44(12): 4787–4792. PubMed Abstract | Publisher Full Text\n\nAl-Omar MA, Amr Ael-G, Al-Salahi RA: Anti-inflammatory, analgesic, anticonvulsant and antiparkinsonian activities of some pyridine derivatives using 2,6-disubstituted isonicotinic acid hydrazides. Arch Pharm (Weinheim). 2010; 343(11–12): 648–56. PubMed Abstract | Publisher Full Text\n\nMeyer EA, Castellano RK, Diederich F: Interactions with aromatic rings in chemical and biological recognition. Angew Chem Int Ed Engl. 2003; 42(11): 1210–50. PubMed Abstract | Publisher Full Text\n\nSalonen LM, Ellermann M, Diederich F: Aromatic rings in chemical and biological recognition: energetics and structures. Angew Chem Int Ed Engl. 2011; 50(21): 4808–42. PubMed Abstract | Publisher Full Text\n\nWaters M: The Role of Aromatic Interactions in Biomolecular Recognition: Contributions to Affinity and Specificity. In Proceedings of the International Beilstein Workshop on Molecular Interactions - Bringing Chemistry to Life. (eds. Hicks M. & Kettner C.). Logos Verlag Berlin, 2006; 111–127. Reference Source\n\nHodnett EM, Dunn WJ 3rd: Structure-antitumor activity correlation of some Schiff bases. J Med Chem. 1970; 13(4): 768–70. PubMed Abstract | Publisher Full Text\n\nHodnett EM, Mooney PD: Antitumor activities of some Schiff bases. J Med Chem. 1970; 13(4): 786. PubMed Abstract | Publisher Full Text\n\nChandraleka S, Ramya K, Chandramohan G, et al.: Antimicrobial mechanism of copper (II) 1,10-phenanthroline and 2,2′-bipyridyl complex on bacterial and fungal pathogens. J Saudi Chem Soc. 2014; 18(6): 953–962. Publisher Full Text\n\nTobriya SK: Biological Applications of Schiff base and its Metal Complexes-A Review. Int J Sci Res. 2014; 3(9): 1254–1256. Reference Source\n\nAl-Garawi ZSM, Tomi IHR, Al-Daraji AHR: Synthesis and Characterization of New Amino Acid-Schiff Bases and Studies their Effects on the Activity of ACP, PAP and NPA Enzymes (In Vitro). J Chem. 2012; 9(2): 962-969. Publisher Full Text\n\nSorenson JRJ, Kishore V, Pezeshk A, et al.: Copper complexes: a physiological approach to the treatment of ‘inflammatory diseases’. Inorganica Chim Acta. 1984; 91(4): 285–294. Publisher Full Text\n\nDaniel KG, Chen D, Yan B, et al.: Copper-binding compounds as proteasome inhibitors and apoptosis inducers in human cancer. Front Biosci. 2007; 12: 135–44. PubMed Abstract | Publisher Full Text\n\nGiovagnini L, Sitran S, Montopoli M, et al.: Chemical and biological profiles of novel copper(II) complexes containing S-donor ligands for the treatment of cancer. Inorg Chem. 2008; 47(14): 6336–43. PubMed Abstract | Publisher Full Text\n\nSorenson JR: Cu, Fe, Mn, and Zn chelates offer a medicinal chemistry approach to overcoming radiation injury. Curr Med Chem. 2002; 9(6): 639–62. PubMed Abstract | Publisher Full Text\n\nSuksrichavalit T, Prachayasittikul S, Piacham T, et al.: Copper complexes of nicotinic-aromatic carboxylic acids as superoxide dismutase mimetics. Molecules. 2008; 13(12): 3040–56. PubMed Abstract | Publisher Full Text\n\nSuksrichavalit T, Prachayasittikul S, Nantasenamat C, et al.: Copper complexes of pyridine derivatives with superoxide scavenging and antimicrobial activities. Eur J Med Chem. 2009; 44(8): 3259–3265. PubMed Abstract | Publisher Full Text\n\nThe National Toxicology Program (NTP) Interagency Center for the Evaluation of & Alternative Toxicological Methods (NICEATM): Test Method Protocol for Solubility Determination Phase III. 2003. Reference Source\n\nStrober W: Trypan blue exclusion test of cell viability. Curr Protoc Immunol. 2001; Appendix 3: Appendix 3B. PubMed Abstract | Publisher Full Text\n\nRepetto G, del Peso A, Zurita JL: Neutral red uptake assay for the estimation of cell viability/cytotoxicity. Nat Protoc. 2008; 3(7): 1125–31. PubMed Abstract | Publisher Full Text\n\nWind M: Current ICCVAM Recommendations for the Use of In Vitro Test Methods to Estimate Acute Systemic Toxicity. In Acute Chemical Safety Testing: Advancing In Vitro Approaches and Humane Endpoints for Systemic Toxicity Evaluations. 2008; 1–21. Reference Source\n\nWorld Health Organization: The WHO recommended classification of pesticides by hazard and guidelines to classification. 2009. Reference Source\n\nWalum E: Acute oral toxicity. Environ Health Perspect. 1998; 106(Suppl 2): 497–503. PubMed Abstract | Free Full Text\n\nKril A, Topashka-Ancheva M, Iliev I, et al.: In vitro antitumour activity, genotoxicity, and antiproliferative effects of aminophosphonic acid diesters and their synthetic precursors. Z Naturforsch C. 2012; 67(9–10): 473–80. PubMed Abstract | Publisher Full Text\n\nChhajed M, Shrivastava AK, Taile V: Synthesis of 5-arylidine amino-1,3,4-thiadiazol-2-[(N-substituted benzyol)]sulphonamides endowed with potent antioxidants and anticancer activity induces growth inhibition in HEK293, BT474 and NCI-H226 cells. Med Chem Res. 2014; 23: 3049–3064. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParasuraman S: Toxicological screening. J Pharmacol Pharmacother. 2011; 2(2): 74–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalakyan M, Babayan N, Grigoryan R, et al.: Dataset 1 in: Synthesis, characterization and toxicity studies of pyridinecarboxaldehydes and aromatic amino acids derived Schiff bases and corresponding copper (II) complexes. F1000Research. 2016. Data Source"
}
|
[
{
"id": "16481",
"date": "21 Sep 2016",
"name": "Carmel E. Mothersill",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI believe that the title is appropriate for the content of this article and the abstract represents an excellent summary of the work. This very interesting paper describes the synthesis and cytotoxicity testing of several Schiff bases with potential use in cancer therapy. The bases are characterised and then tested for toxicity to human cell lines using two well established tests. The results reveal differences in responses of the two cell lines (one blood and one epithhelial derived) making them potentially very useful. The authors also hint at different biological mechanisms being involved.\n\nThe structure, methods and analysis of the results from the study are very well designed. I think the conclusions are sensible, balanced and justified on the basis of the results of the study. There is enough information provided to be able to replicate the experiment and the data are in a usable format/ structure and have all been provided.\n\nIn summary, the paper is well written and the data appear sound.",
"responses": []
},
{
"id": "17256",
"date": "04 Nov 2016",
"name": "Salvatore Failla",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe focuses of the paper of Malakyan, Arakelyan and co-workers are the synthesis and characterization of isomeric 2-, 3- and 4-pyridinecarboxaldehydes and L-tryptophan derived Schiff bases and their copper (II) complexes and the study of their cytotoxic activity respect to two different cell lines, in particular the Human HeLa (cervix carcinoma) and the KCL-22 (chronic myeloid leukemia).\n\nThe paper is well written, clear and highlights the differences between the cytotoxic activity of all Schiff base ligands and relative copper (II) complexes investigated through toxicity profiles. Moreover, the characterization of all compounds involved is detailed described and, in general, we find the results achieved from this investigation interesting.\n\nWe suggest discussing two points to improve the quality of paper:\nHow metal complexation influences the cytotoxic activity of Cu-2pyr.Trp, Cu-3pyr.Trp and Cu-4pyr.Trp respect to relative Schiff base ligands.\n\nThe effect of the localization of carboxaldehyde group at 2-, 3- or 4-position with regard to nitrogen of aromatic ring on cytotoxic activity of all compound investigated.\n\nFinally, a minor revision. In the first paragraph of the Results and Discussion section, especially in the “Synthesis and characterization of Schiff bases and their copper (II) complexes” section, the sentence “Based on 1H NMR spectra inspection for synthesized Schiff bases in D2O and CD3OD, the loss of 2 singlet signals of CH2 group was revealed, confirming the presence of target compounds.” seems in contrast with the next one. This sentence should be reworked considering that the CH=N 1H NMR signal is diagnostic for evaluation of the Schiff bases formation.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1921
|
https://f1000research.com/articles/5-1920/v1
|
05 Aug 16
|
{
"type": "Research Article",
"title": "Ability of device to collect bacteria from cough aerosols generated by adults with cystic fibrosis",
"authors": [
"David N. Ku",
"Sarah K. Ku",
"Beth Helfman",
"Nael A. McCarty",
"Bernard J. Wolff",
"Jonas M. Winchell",
"Larry J. Anderson",
"Sarah K. Ku",
"Beth Helfman",
"Nael A. McCarty",
"Bernard J. Wolff",
"Jonas M. Winchell",
"Larry J. Anderson"
],
"abstract": "Background: Identifying lung pathogens and acute spikes in lung counts remain a challenge in the treatment of patients with cystic fibrosis (CF). Bacteria from the deep lung may be sampled from aerosols produced during coughing. Methods: A new device was used to collect and measure bacteria levels from cough aerosols of patients with CF. Sputum and oral specimens were also collected and measured for comparison. Pseudomonas aeruginosa, Staphylococcus aureus, Klebsiella pneumoniae, and Streptococcus mitis were detected in specimens using Real-Time Polymerase Chain Reaction (RT-PCR) molecular assays. Results: Twenty adult patients with CF and 10 healthy controls participated. CF related bacteria (CFRB) were detected in 13/20 (65%) cough specimens versus 15/15 (100%) sputum specimens. Commensal S. mitis was present in 0/17 (0%, p=0.0002) cough specimens and 13/14 (93%) sputum samples. In normal controls, no bacteria were collected in cough specimens but 4/10 (40%) oral specimens were positive for CFRB. Conclusions: Non-invasive cough aerosol collection may detect lower respiratory pathogens in CF patients, with similar specificity and sensitivity to rates detected by BAL, without contamination by oral CFRB or commensal bacteria.",
"keywords": [
"Cystic fibrosis",
"aerosols",
"specimen collection",
"respiratory infections",
"etiology"
],
"content": "Introduction\n\nThe etiology of lower respiratory tract infections in the lungs is difficult to determine, in part because a good quality specimen from the site of the infection is not readily available1–4. Access to such a specimen would be an important advance in the monitoring and treatment of cystic fibrosis (CF), as well as other lower respiratory tract infections, such as pneumonia, tuberculosis, asthma, lung cancer, etc. Presently, oropharyngeal (OP), sputum, and bronchoalveolar lavage (BAL) specimens are typically used to monitor CF patients. OP specimens may be appropriate for detecting viruses, but are not ideal for most bacterial pathogens. Sputum is commonly collected to monitor CF but often contains contaminants and cystic fibrosis related bacteria (CFRB) from the upper respiratory tract. The difficulty some patients have in producing an acceptable sputum specimen further decreases the value of these samples, often causing the physician to treat the patient empirically1–4. BAL provides a specimen from the lungs but is an invasive procedure that cannot be routinely used. BAL specimens may also collect contaminants from the upper respiratory tract5–7.\n\nAn alternative source for a lung specimen is from aerosols generated during coughs8–12. Studies show that one cough can generate as many as 66,000 expelled particles10,13. Patients that have lower respiratory tract infections can infect others through respiratory dispersion of pathogens in aerosols generated by coughing or sneezing. Coughing produces a higher concentration of pathogens from the lower lungs than normal exhalation or sneezing8–13. A new cough specimen collection device (PneumoniaCheck™, Figure 1) collects aerosols from the lungs onto a micropore filter while minimizing contamination from the upper respiratory tract. Microbiology or molecular assays can then be used to detect pathogens collected on the device’s filter.\n\nThe device uses a reservoir to separate oral contents from deep lung aerosols using fluid mechanics for separation (Figure 2). The initial volume of air that comes from exhalation or coughing is contaminated air from the upper respiratory tract, also known as anatomic dead space. When a patient coughs into the device, this air from the upper airways first flows into the reservoir (Figure 2a). The exhaled air flows to the reservoir first as it has the least resistance compared to the filter at the end of the device. This reservoir has a volume of 250 ml, approximately 100 ml greater than the volume of anatomic dead space in the average adult15, which ensures that all of the upper airway aerosols are completely separated out. The expanded reservoir is inelastic, creating a back-pressure, so subsequent exhaled breath is forced through the microbial filter (Figure 2b). Therefore, only lung aerosol contents are collected onto the filter and are free from upper airway contamination.\n\n(a) Contaminated upper airway particles from the mouth initially fill up air reservoir. (b) Then, uncontaminated lower airway particles from the lungs are captured onto filter.\n\nA previous study demonstrated that the device’s filter is >99% effective in collecting airborne bacteria (approximately 3.1 μm in diameter) and viruses (approximately 2.8 μm in diameter)16. Sampling from normal individual controls showed zero collection of oral contents on the filter, even with up to 15 ml of liquid in the mouth (simulating sputum). The PneumoniaCheck™ device has been shown to significantly separate the lower airway gas from the upper airway gas based on oxygen and alcohol levels (p<0.0001)16.\n\nCF is a genetic disease that affects the lungs of approximately 28,000 children and adults in the United States each year17. People with CF often have chronic lung infections and require regular monitoring to ensure that bacterial colonization does not develop into infection26,27. We used specimens from sputum and coughs to compare their abilities to capture, identify, and quantify relative levels of lung bacteria in adult CF patients. The goal of this study is to determine if the cough device can capture lung pathogens from adult patients with chronic lung infection while simultaneously excluding oral bacteria.\n\n\nMaterials and methods\n\nPatients with CF (n=20) aged >18 years old were recruited from the Emory Cystic Fibrosis Center Adult Clinic in Atlanta, Georgia. The Emory Institutional Review Board (H08353) approved the study and participants provided their written, informed consent. The sample size was sufficiently powered to demonstrate statistical significance for lower lung sampling without oral contamination. Tests of paired proportions were conducted using an exact form of the McNemar test to compare the presence of CFRB between two samples (i.e. cough and sputum). The Wilcoxon signed rank test was used to compare cycle threshold (CT) values between the different methods of sampling. The CT value of 60 was used as the upper limit of detection for all PCR assays to determine relative quantity of bacteria in each specimen.\n\nThroat swabs and cough device specimens were collected from 10 healthy, non-smoking subjects for normal controls. Separately, a sputum specimen and cough device specimen were each collected from 20 adult patients with CF. Cough device specimen collection preceded sputum specimen collection in order to help induce sputum. Specimen collections were supervised and emergency equipment was readily available. Streptococcus mitis is a commensal bacterium that is found in the mouth but not in the lungs14. Streptococcus pneumoniae and Staphylococcus aureus are also commonly found in the oral cavity3. Pseudomonas aeruginosa, Staphylococcus aureus and Klebsiella pneumoniae are cystic fibrosis related bacteria (CFRB)18–20. Oral and cough specimens were analyzed for these bacteria to determine levels of oral contamination.\n\nFennelly’s Cough Aerosol Sampling System (CASS)29,30 and Knibbs’ Distance Rig13 have demonstrated that cough particles can carry substantial concentrations of bacteria from lower respiratory infections. A previous article on the cough collection device describes the ability of the device to selectively sample from the lower lungs while excluding oral contaminants16. The cough device used in this study provides a less cumbersome option to Fennelly’s and Knibbs’ methods for lung specimen collection. Each patient coughed 10 times into the device to ensure sufficient aerosol collection.\n\nMicrobiology culturing has several limitations that decrease the efficiency and effectiveness of rapid diagnosis24. Throat, sputum, and cough specimens were all analyzed using molecular PCR methods. All specimens were processed in a BSL 2 safety cabinet. The cough device filter was removed, placed into a 2 ml sterile freezer vial, and stored at -80°C. Respiratory secretions captured on the filter were removed by hydrating the filter with 1mL of lysis buffer (MagNA Pure LC lysis buffer; Roche Applied Science, Indianapolis, IN), vortexing, incubating for 5 min at room temperature, and collected using a pipette. Fluid remaining in the filter was collected by placing the filter in a sterile Costar SpinX microfuge tube with a 0.45 micron filter (Corning Inc., Corning, NY), centrifuging for 1 min at 10,000 rpm, and retrieved using a pipette. The residual fluid was then combined with original collected fluid and then 400 μL was extracted on the MagNA Pure Compact Instrument (Roche Applied Science) per the manufacturer’s instructions. The extracted nucleic acid was eluted into 100 μL of elution buffer and stored at -80°C for qPCR testing.\n\nThe sputum specimen was mixed with 1 mL of phosphate buffered saline (PBS), homogenized with pipetting and vortexing, mixed with a 12.5 mM equal volume of freshly prepared dithiothreitol (DTT, No Weigh™ format, Fisher Scientific), and incubated at room temperature for 30 min with periodic vortexing. The resultant solution was divided into 400 μL aliquots and stored at -80°C. A 400 μL aliquot of the processed sample was then extracted on the MagNA Pure Compact Instrument and stored as described above.\n\nThe extracted nucleic acid was tested for P. aeruginosa, S. aureus, K. pneumoniae, and S. mitis targets by individual real-time PCR assays. The primer and probe sequences for these assays have been previously described25. The S. mitis primers are: Forward TTTTGTCATCTAGCCTTGC; Reverse GCAGTCATATCATCACCTTC and Probe ACTTGGGCAATCCCGACAGATTCTAAC, with a 5' FAM reporter and a 3' BHQ quencher. The PCR reactions were done with 5 μl of extracted nucleic acid from the specimens plus 12.5 μl of PerfeCTa Multiplex qPCR SuperMix (catalog no. 95063-200; Quanta BioSciences), 0.5 μM final concentrations of each primer, 0.1 μM final concentration of the probe, and nuclease-free water (catalog no. P1193; Promega) to a final reaction volume of 25 μL. Real-time PCR reactions were performed using an ABI 7500 standard machine (Life Technologies, Carlsbad, CA) with enzyme activation at 95°C for 5 min, followed by 45 cycles of 95°C for 15 seconds and 60°C for 1 min. All specimens from CF patients were run in duplicate for each target. The CT values for individual PCR assays were used as an indication of the relative quantity of bacteria in the specimen.\n\n\nResults\n\nAll subjects completed specimen collection safely. Ten healthy subjects were used for controls. Sputum and cough specimens were successfully collected from 20 adult patients with CF, with the exception of five patients who could not produce a sputum specimen.\n\nNormal controls demonstrated a high incidence of false positives from oral sampling, shown in Table 1. Bacteria were isolated from throat swabs in 4/10 (40%) normal, healthy control subjects. S. pneumoniae was positive in 2/10 (20%) oral specimens and S. aureus was positive in 3/10 (30%) oral specimens, with one subject positive for both bacteria. In contrast, 0/10 (0%, p=0.0313) cough specimens were positive for bacteria in normal controls. The calculated true negative rate or specificity for sputum specimens was 60% and 100% for cough specimens.\n\nSpecificity in the CF patients was similar. For the CF patients, S. mitis was isolated from 13/14 (93%) sputum specimens but in none of the cough specimens (0%, p=0.0002). The cough specimens collected no S. mitis. CFRB was collected in both specimen types. P. aeruginosa was isolated from 13/15 (87%) sputum specimens and 9/20 (45%) cough specimens (p=0.0213). S. aureus was isolated from 9/15 (60%) sputum specimens and 3/20 (15%) cough specimens. K. pneumoniae was isolated from 2/15 (13%) sputum specimens and 3/20 (15%) cough specimens.\n\nIn aggregate, sputum specimens were positive for CFRB in 15/15 (100%) samples. Cough specimens were positive for CFRB in 13/20 (65%) samples. The sputum specimens had a 93% rate of oral commensals. Sputum specimens were positive for three or more pathogens in 2/15 (13%) samples, and positive for two or more pathogens in 7/15 (47%) samples. In contrast, cough specimens had no commensals and were positive in 65% of the CF patients. The cough specimens were positive for two or more pathogens in 2/20 (10%, p<0.05) specimens and no cough specimens were positive for three or more pathogens. The results of these real-time PCR identifications are listed in Table 2.\n\n* Five patients were unable to produce viable sputum specimens\n\n† Six sputum specimens were not tested for S. mitis\n\n§ Three cough specimens were not tested for S. mitis\n\nCT values are inversely proportional to the quantity of bacteria in a sample, i.e. small values indicate higher quantities of colony forming units (CFU). For the CFRB samples, P. aeruginosa CT values ranged from 18–33 in sputum specimens and 33–42 in cough specimens. S. aureus CT values ranged from 24–38 in sputum specimens and 36–40 in cough specimens. For both P. aeruginosa and S. aureus the cough and sputum specimens significantly differed in CT values (p=0.0017 and 0.0092, respectively). K. pneumoniae CT values ranged from 37–38 in sputum specimens and 39–43 in cough specimens. Thus, the CT values for cough specimens were consistently higher than those of sputum. Note that the cough filter samples from normal controls exhibited no pathogens up to CT values of 60.\n\n\nDiscussion\n\nIt is widely recognized that a simple, safe, non-invasive, low maintenance, inexpensive, widely accessible sampler is needed for the collection of lower respiratory pathogens34–37.\n\nCollection of lower lung contents by coughing is much easier than BAL specimen collection. The method is convenient for patients who are already inclined to cough and they reported that the use of the device helped clear their lungs. Cough specimens may provide a non-invasive yet specific sample for in-home surveillance to watch for spikes in lung pathogens in patients with CF. Use of the device to collect cough aerosols has the potential to provide a clean alternative to oral samples for detecting lower lung pathogens.\n\nAs is well known, commonly used samples of sputum or OP swabs show strong contamination in the upper airway3,4. Nearly all of the sputum specimens were positive for the oral commensal S. mitis. In contrast, none of the cough specimens were positive for S. mitis. This difference in the commensal S. mitis between sputum and cough specimens reiterates the unreliability and low specificity of sputum3,4. Further, not all patients can produce an adequate sputum sample. Examining the 12 subjects with paired sputum and S. mitis results, 8/12 (67%) sputum and cough specimens had concordant positives, although six of these eight were positive for additional bacteria in sputum. These six sputum specimens with multiple bacteria likely indicate false positives from oral contamination rather than co-infection.\n\nThe number of concordant sputum and cough specimens (2/12, 17%) was small. Conversely, the pathogen detected in the cough specimen was different from that observed in the sputum specimen. 2/12 (17%) sputum specimens did not identify bacteria that were identified in cough specimens, possibly reflecting masked readings associated with commensal distraction. The differences demonstrate that the cough device is not just collecting sputum.\n\nP. aeruginosa is the most common bacterium found in lungs of adult CF patients19, and was also the most prevalent bacterium collected in our cohort. 13/20 (65%) cough specimens were positive for CFRB. This incidence and distribution of pathogens in CF is similar to the 59% positive for CFRB in BAL sampling22,23. Prior series of BAL specimens in similar populations have yielded positive CFRB of 59–85%, similar to the 65% positivity from the cough device illustrating comparable sensitivity22,23.\n\nCollection of exhaled aerosols has been studied by several previous groups. An alternate device for aerosol collection is the RTube™; however, it varies greatly in design and function28. The RTube™ system is designed to collect from all exhaled breath that condenses28 while PneumoniaCheck™ is designed to collect particulate sized lung aerosols and separate out the mouth contents16. The majority of exhaled gas passes out of the end of the RTube™ as only water condensate is intended to be collected. Exhaled breath condensate can be a useful specimen for identifying pH levels, but is generally not viewed as a reliable specimen for identifying lower respiratory infections34–37.\n\nWainwright et al., reported detecting P. aeruginosa in cough aerosols by culture12. They reported 25/28 (89%) positive in a mixed population of children and adults with CF using a cough aerosol sampling system (CASS) for 5 minutes with each subject12. Similarly, Knibbs et al. reported that 14/18 (78%) patients aerosolized P. aeruginosa that remained viable and presumably transmissive up to 45 minutes after coughs sampled on an Anderson impactor13. Knibbs et al. used conventional microbiology cultures to quantify colony forming units. Both of these studies used a specially constructed aerosol sampler that is expensive, cumbersome, and difficult to use in a clinical setting. For these studies, patients cough into a standard mouthpiece and aerosols are sucked into impactors using vacuum air pumps. The CASS system was not designed for routine use in clinical settings and the mouthpiece was not designed to exclude oral contents. These designs differ from PneumoniaCheck™, which has a mouthpiece designed to specifically exclude oral contaminants16. The high incidence of P. aeruginosa, using the snorkel type mouthpiece and tubing, may reflect some collection of oral contents using CASS.\n\nRT-PCR may be used to quantify the amount of pathogens in a sample. As more material is collected on the filter, the CT counts will drop similar to the inverse of CFUs32. It should be noted that an aerosolized lung specimen should have higher CT values compared to the liquid specimens of sputum due to a lack of contamination and the small physical volume of aerosols. CT values in the sputum specimens ranged from 19–38, whereas the range in cough specimens was 33–43 (p<0.001, Table 3). Nonetheless, CT values in all positive cough specimens are significantly lower than the baseline of >60 for normal controls. While the CT values are higher for the cough specimens, the background noise level of the virgin filter is >60, allowing limits of detection by PCR that may be more sensitive to lung CFRB.\n\nOne can utilize the CT values to compare relative amounts of pathogens being coughed by an individual patient compared with a population32,33. Jones-López et al. found that both CF and TB patients can produce aerosols with viable pathogens, but the amount of pathogens produced by individuals varies greatly31. Patients with high amounts of M. tuberculosis in cough aerosols were more likely to have transmitted to others30. Those that produce large amounts of pathogens in coughs may be more efficient transmitters, e.g. “superspreaders”13,30. The quantity of pathogens in a cough may be a critical metric in transmission of infectious disease, controlling epidemics, and monitoring colonization. In the Jones-López et al. study, the amount of aerosolized M. tuberculosis fell dramatically after three weeks of treatment. Therefore, a cough specimen could also be used to monitor levels of resistant bacteria, if present.\n\nPublished guidelines for CF patients suggest acquiring quarterly respiratory specimens to monitor lung infections26,27. Cough specimens may be a more specific and sensitive method for monitoring colonization and determining infectivity. Additional studies could explore this application further by comparing CT values with symptoms. If a CF patient is monitored on a regular basis using cough specimens, a sudden decrease in CT value may indicate a change in pathogen burden32,33. The CT value of pathogen burden in cough aerosols may be useful as a measurement to determine if the lung burden is growing.\n\nThis study has several limitations. Our study included only adult patients; thus, we cannot comment on the aerosol production during coughing by pediatric patients. This study reports on 20 patients. Most studies in the literature have a similar number of subjects, since lower respiratory identification has always been an enormous challenge12,13. Future studies may evaluate the benefits of requiring more coughs or coughing for a specified amount of time, such as 5 minutes, to establish CT thresholds for this new method of specimen collection. The RT-PCR molecular assays used in this study are not available at all hospitals, although a few commercial laboratories can provide clinical respiratory identification services.\n\n\nConclusion\n\nIn summary, we have shown that a new device can collect lung pathogens from adult patients with CF from cough aerosols with identification using molecular assays. The device excludes oral contaminants showing higher specificity than sputum samples. Identifying causative pathogens in the lower respiratory tract is likely to play a significant role in patient management24. The data in this study suggest an alternative to sputum collection for the identification of lower respiratory pathogens.\n\n\nConsent\n\nWritten informed consent was obtained by all participants through Institutional Review Board Protocol #000-2492 approved by Georgia Institute of Technology, Emory University, and US Centers for Disease Control and Prevention.\n\n\nData availability\n\nAll raw data are provided in the tables above.",
"appendix": "Author contributions\n\n\n\nDavid N. Ku, Nael A. McCarty, Bernard J. Wolff, Jonas M. Winchell, and Larry J. Anderson served as scientific advisors. Beth Helfman collected data and provided and cared for study patients. Sarah K. Ku wrote and edited much of the manuscript. All authors agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nDr. David N. Ku, Sarah K. Ku, and Dr. Larry J. Anderson are co-inventors on the PneumoniaCheck™ patent licensed to MD Innovate, Inc. from US Centers for Disease Control and Prevention and Georgia Tech Research Corporation.\n\n\nGrant information\n\nSupported by internal grants from US Centers for Disease Control and Prevention and Emory University Hospital.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBurns MW: Precipitins to Klebsiella and Other Enterobacteria in the Serum of Patients with Chronic Respiratory Disorders. The Lancet. 1968; 1(7539): 383–385. PubMed Abstract | Publisher Full Text\n\nGoddard AF, Staudinger BJ, Dowd SE, et al.: Direct sampling of cystic fibrosis lungs indicates that DNA-based analyses of upper-airway specimens can misrepresent lung microbiota. Proc Natl Acad Sci U S A. 2012; 109(34): 13769–13774. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLentino JR, Lucks DA: Nonvalue of Sputum Culture in the Management of Lower Respiratory Tract Infections. J Clin Microbiol. 1987; 25(5): 758–762. PubMed Abstract | Free Full Text\n\nRosenfeld M, Emerson J, Accurso F, et al.: Diagnostic accuracy of oropharyngeal cultures in infants and young children with cystic fibrosis. Pediatr Pulmonol. 1999; 28(5): 321–328. PubMed Abstract | Publisher Full Text\n\nGilchrist FJ, Salamat S, Clayton S, et al.: Bronchoalveolar lavage in children with cystic fibrosis: how many lobes should be sampled? Arch Dis Child. 2011; 96: 215–217. PubMed Abstract | Publisher Full Text\n\nWillner D, Haynes MR, Furlan M, et al.: Spatial distribution of microbial communities in the cystic fibrosis lung. ISME J. 2012; 6(2): 471–474. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWimberley N, Faling LJ, Bartlett JG: A Fiberoptic Bronchoscopy Technique to Obtain Uncontaminated Lower Airway Secretions for Bacterial Culture. Am Rev Respir Dis. 1979; 119(3): 337–343. PubMed Abstract\n\nCouch RB, Cate TR, Douglas RG Jr, et al.: Effect of Route of Inoculation on Experimental Respiratory Viral Disease in Volunteers and Evidence for Airborne Transmission. Bacteriol Rev. 1966; 30(3): 517–529. PubMed Abstract | Free Full Text\n\nCouch RB, Cate TR, Gerone PJ, et al.: Production of Illness with a Small-Particle Aerosol of Coxsackie A21. J Clin Invest. 1965; 44(4): 535–542. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGerone PJ, Couch RB, Keefer GV, et al.: Assessment of Experimental and Natural Viral Aerosols. Bacteriol Rev. 1966; 30(3): 576–584. PubMed Abstract | Free Full Text\n\nLee N, Chan PK, Hui DS, et al.: Viral loads and duration of viral shedding in adult patients hospitalized with influenza. J Infect Dis. 2009; 200(4): 492–500. PubMed Abstract | Publisher Full Text\n\nWainwright CE, France MW, O’Rourke P, et al.: Cough-generated aerosols of Pseudomonas aeruginosa and other Gram-negative bacteria from patients with cystic fibrosis. Thorax. 2009; 64(11): 926–931. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnibbs LD, Johnson GR, Kidd TJ, et al.: Viability of Pseudomonas aeruginosa in cough aerosols generated by persons with cystic fibrosis. Thorax. 2014; 69(8): 740–745. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiljemark WF, Gibbons RJ: Proportional distribution and relative adherence of streptococcus miteor (mitis) on various surfaces in the human oral cavity. Infect Immun. 1972; 6(5): 852–859. PubMed Abstract | Free Full Text\n\nHart MC, Orzalesi MM, Cook CD: Relation between anatomic respiratory dead space and body size and lung volume. J Appl Physiol. 1963; 18(3): 519–522. Reference Source\n\nScholz TL, Midha PA, Anderson LJ, et al.: PneumoniaCheck: A Device for Sampling Lower Airway Aerosols. J Med Devices. 2010; 4(4): 041005. Publisher Full Text\n\nCystic Fibrosis Foundation Patient Registry: 2013 Annual Data Report to the Center Directors. Bethesda, Maryland. 2014.\n\nFurness JC, Habeb A, Spencer DA, et al.: To the editor: Bronchoalveolar lavage (BAL) in pediatric cystic fibrosis (CF): its clinical use modified by audit in a regional CF center. Pediatr Pulmonol. 2002; 33(3): 234. PubMed Abstract | Publisher Full Text\n\nLyczak JB, Cannon CL, Pier GB: Lung infections associated with cystic fibrosis. Clin Microbiol Rev. 2002; 15(2): 194–222. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamsey BW, Wentz KR, Smith AL, et al.: Predictive value of oropharyngeal cultures for identifying lower airway bacteria in cystic fibrosis patients. Am Rev Respir Dis. 1991; 144(2): 331–337. PubMed Abstract | Publisher Full Text\n\nArmstrong DS, Grimwood K, Carlin JB, et al.: Bronchoalveolar lavage or oropharyngeal cultures to identify lower respiratory pathogens in infants with cystic fibrosis. Pediatr Pulmonol. 1996; 21(5): 267–275. PubMed Abstract | Publisher Full Text\n\nHarris JK, De Groote MA, Sagel SD, et al.: Molecular identification of bacteria in bronchoalveolar lavage fluid from children with cystic fibrosis. Proc Natl Acad Sci U S A. 2007; 104(51): 20529–20533. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZemanick ET, Wagner BD, Harris JK, et al.: Pulmonary exacerbations in cystic fibrosis with negative bacterial cultures. Pediatr Pulmonol. 2010; 45(6): 569–577. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOosterheert JJ, Bonten MJ, Buskens E, et al.: Algorithm to determine cost savings of targeting antimicrobial therapy based on results of rapid diagnostic testing. J Clin Microbiol. 2003; 41(10): 4708–4713. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEmery SL, Erdman DD, Bowen MD, et al.: Real-time reverse transcription-polymerase chain reaction assay for SARS-associated coronavirus. Emer Infect Dis. 2004; 10(2): 311–316. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaiman L, Siegel J, Cystic Fibrosis Foundation: Infection control recommendations for patients with cystic fibrosis: microbiology, important pathogens, and infection control practices to prevent patient-to-patient transmission. Infect Control Hosp Epidemiol. 2003; 24(5 Suppl): S6–52. PubMed Abstract | Publisher Full Text\n\nYankaskas JR, Marshall BC, Sufian B, et al.: Cystic fibrosis adult care: consensus conference report. Chest. 2004; 125(1 Suppl): 1S–39S. PubMed Abstract | Publisher Full Text\n\nSoyer OU, Dizdar EA, Keskin O, et al.: Comparison of two methods for exhaled breath condensate collection. Allergy. 2006; 61(8): 1016–1018. PubMed Abstract | Publisher Full Text\n\nFennelly KP, Martyny JW, Fulton KE, et al.: Cough-generated aerosols of Mycobacterium tuberculosis: a new method to study infectiousness. Am J Respir Crit Care Med. 2004; 169(5): 604–609. PubMed Abstract | Publisher Full Text\n\nFennelly KP, Jones-López EC, Ayakaka I, et al.: Variability of infectious aerosols produced during coughing by patients with pulmonary tuberculosis. Am J Respir Crit Care Med. 2012; 186(5): 450–457. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJones-López EC, Namugga O, Mumbowa F, et al.: Cough aerosols of Mycobacterium tuberculosis predict new infection: a household contact study. Am J Respir Crit Care Med. 2013; 187(9): 1007–1015. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMehta T, McGrath E, Bheemreddy S, et al.: Detection of oseltamivir resistance during treatment of 2009 H1N1 influenza virus infection in immunocompromised patients: utility of cycle threshold values of qualitative real-time reverse transcriptase PCR. J Clin Microbiol. 2010; 48(11): 4326–4328. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEspy MJ, Uhl JR, Sloan LM, et al.: Real-time PCR in clinical microbiology: applications for routine laboratory testing. Clin Microbiol Rev. 2006; 19(1): 165–256. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu Z, Shen F, Li X, et al.: Molecular and microscopic analysis of bacteria and viruses in exhaled breath collected using a simple impaction and condensing method. PLoS One. 2012; 7(7): e41137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHouspie L, De Coster S, Keyaerts E, et al.: Exhaled breath condensate sampling is not a new method for detection of respiratory viruses. Virol J. 2011; 8: 98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFabian P, McDevitt JJ, DeHaan WH, et al.: Influenza virus in human exhaled breath: an observational study. PLoS One. 2008; 3(7): e2691. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuynh KN, Oliver BG, Stelzer S, et al.: A new method for sampling and detection of exhaled respiratory virus aerosols. Clin Infect Dis. 2008; 46(1): 93–95. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "16100",
"date": "21 Sep 2016",
"name": "Mats Kalin",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nKu et al, describe the experiences with a newly designed device for collecting cough specimens from patients with CF. The device is constructed with the intention that dead space air, presumably containing a high concentration of oral contaminants, should be collected separately, while on the other hand cough material from the lower lungs is to be collected on a specific filter constructed so that bacteria should be trapped in such a way that the material may be used for RT-PCR.\nTitle, abstract, methods and material are clearly described as is results. Discussion is adequate and relevant.\nThe presented results indicate high specificity with low risk of oral contaminants in cough specimens than in sputum from 20 adult CF individuals. Actually only 12 patients produced a sputum specimen, so it is a small study. However the differences were significant.\nSensitivity cannot be assessed with the way the study was carried out, but comparison with other studies indicate satisfactory results. Thus, the device seems to permit improved analysis of quantitative bacteriology in lower lung specimens from CF patients. The device is described as simple to use and is suggested to be used to follow lower respiratory tract microbiology in CF patients, so that increased concentrations of significant bacterial pathogens may be noted. This may be a step forward for the management of these patients. Further studies, including pediatric studies, are needed to corroborate the findings in this study and to explore the advantages of longitudinal follow up of CF patients\nPCR may not suffice all the time, since bacterial resistance may have to be detected and specified in order to find an adequate treatment alternative. Needless to say this is an increasing problem.",
"responses": []
},
{
"id": "16745",
"date": "03 Oct 2016",
"name": "David E Griffith",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nComment #1: I am not familiar with the term Cycle Threshold, “CT”. The term is introduced in the first paragraph of the Methods and then mentioned again in the last paragraph of the Methods. It would have been helpful to me to have had a brief discussion of this term in the Introduction or Methods. The CT values were discussed in full paragraphs in the Results and Discussion sections so it is clearly an important metric for this study and there could be other readers not familiar with CT.\n\nComment #2: Evaluation of microbiologic results in bronchiectasis patients is difficult. First, there is no “gold standard” test for identifying potential respiratory pathogens in the lungs of these patients. Microbiome studies suggest that a large number of potential respiratory pathogens populate the lungs of these patients but do not help determine which one(s) are responsible for symptoms or clinical deterioration and therefore might benefit from therapy. Similarly, microbiology results from BAL do not necessarily identify pathogens responsible for symptoms and clinical deterioration. I think these observations are pertinent with regard to the sensitivity and specificity claims of the authors. In Table 1, it is apparent that mouth flora is not sampled with the cough technique avoiding an important mechanism of specimen contamination. In Table 2, however, there is poor concordance between sputum and cough with regard to Pseudomonas and Staph, with these 2 potential pathogens isolated more commonly with sputum than cough. One interpretation of that observation is that sputum is more sensitive than cough for recovering potential respiratory pathogens in CF. The more frequent isolation of multiple respiratory pathogens with sputum could be interpreted the same way especially in light of the microbiome data. I am unsure where the authors believe the source of the “excess” Pseudomonas and Staph (as well as the specimens with multiple pathogens) is for sputum patients? Are they suggesting these “excess” isolates are “contaminants”?\n\nComment #3: I think the authors have convincingly shown that the PneumoniaCheck™ device avoids upper airway bacterial contamination during specimen collection with cough in CF patients. I think their findings with regard to the number and type of CF related respiratory pathogens and the clinical significance of those pathogens remains to be elucidated.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1920
|
https://f1000research.com/articles/5-1915/v1
|
04 Aug 16
|
{
"type": "Opinion Article",
"title": "Zika virus emergency in Brazil: scientific challenges and early developments",
"authors": [
"Tazio Vanni",
"Karlos Diogo Chalegre",
"Camile Giaretta Sachetti",
"Pedro Reginaldo Prata",
"Marco Antônio Fireman",
"Karlos Diogo Chalegre",
"Camile Giaretta Sachetti",
"Pedro Reginaldo Prata",
"Marco Antônio Fireman"
],
"abstract": "The epidemic of microcephaly and other congenital abnormalities associated with Zika virus which emerged in Brazil now threatens different countries worldwide. Since the declaration of a National Public Health Emergency, the Brazilian government has implemented a response plan in which the research agenda is central. Developments were achieved in four main areas of the agenda: 1) virological, clinical and epidemiological studies, 2) alternative vector control strategies, 3) development and evaluation of diagnostic tests, and 4) development and evaluation of vaccines. National and international collaborative networks have played an important role in the race against the clock to quickly translate the results of R&D initiatives into public policies. It is paramount that the lessons learned from Zika lead to fast and effective responses to future epidemics.",
"keywords": [
"Zika virus",
"epidemic",
"microcephaly",
"Aedes Aegypti",
"vector control",
"scientific agenda"
],
"content": "\n\nIn the last decades, Brazil has faced different arbovirus epidemics. However, none of them had the complexity of Zika virus and associated diseases. In April 2015, the first cases of the virus were reported in the country1. Initially, the occurrence was considered to be of no greater threat than dengue or chikungunya. Nonetheless, by the end of October, the number of microcephaly cases started to rise sharply, which triggered a thorough investigation and subsequently the declaration of a National Public Health Emergency2,3. On December 5, the President of Brazil launched the National Microcephaly Response Plan, involving 19 institutions and structured on three pillars: 1) vector control, 2) health care, and 3) research & education2.\n\nThe research agenda focused on four main areas: 1) virological, clinical and epidemiological studies, 2) alternative vector control strategies, 3) development and evaluation of diagnostic tests, and 4) development and evaluation of vaccines. After almost one year since the reporting of the first cases of microcephaly associated with Zika in the country, many developments in the agenda were achieved and other challenges emerged:\n\n1) Virological, clinical, and epidemiological studies – Researchers in Brazil were able to characterize transplacental Zika transmission and its influence in halting neurological development4,5. These findings supported campaigns to increase awareness and protection of pregnant women against mosquitoes. Although Zika seems to be the main culprit of microcephaly increase, other cofactors are under investigation, what may lead to new policies to tackle other risk factors6. Recent studies also suggest that the consequences of Zika infection go beyond microcephaly, pointing out the need to further characterize syndromes and related diseases as well as to revise diagnostic and management protocols7.\n\n2) Alternative vector control strategies – After the Zika emergency was declared, a range of new vector control strategies were proposed, which target different phases of the mosquito life cycle and different settings. The Brazilian Ministry of Health has been promoting effectiveness evaluations of promising strategies, including Wolbachia-infected mosquitoes and mosquito-driven dissemination of pyriproxyfen8,9. These studies will provide invaluable information to improve Aedes control policies in Brazil. Entomological studies have also been investigating if Aedes Aegypti is the only Zika virus vector in Brazil10. This is a crucial point because other mosquito species have different breeding and feeding habits; in which case, the results of these studies may have an important impact on vector control measures.\n\n3) Development and evaluation of diagnostic tests – Since the first cases of Zika have been identified there has been an ongoing effort to improve molecular tests and to develop highly sensitive and specific serological tests, with limited cross-reaction with other arbovirus, allowing point-of-care utilization11. Candidates have arisen from private and public initiatives, which are being validated and evaluated with support from the Brazilian Ministry of Health. The inclusion of such tests in the public health system will require training of health professionals and modifying follow-up protocols. As the spectrum of Zika consequences widens, so does the need for detection and treatment.\n\n4) Development and evaluation of vaccines – The development of an effective and secure vaccine against Zika has been one of the main goals worldwide. Different research groups are working on that, including groups in Brazil. Nonetheless, only one vaccine candidate has received FDA approval to initiate a phase I clinical trial12. Brazilian governmental bodies, such as the National Research Ethics Council, the National Clinical Trials Registry and the National Health Surveillance Agency, developed task-forces to timely evaluate research projects, clinical trials, and products related to Zika virus and associated diseases.\n\nIt has also been a race against the clock to quickly translate the results of R&D initiatives into public policies. For this purpose, the Brazilian Ministry of Health set up the Zika and Related Diseases Specialists Network, fostering greater collaboration between researchers and decision makers13. The joint effort between the Ministries of Health; Science, Technology, Innovation and Communication; and Education also made possible the launching of an open call for strategic research projects to tackle this emergency.\n\nInternational research collaborations were established with partners such as the Center for Diseases Control, the World Health Organization, the US National Institutes of Health and the British Council. Since WHO declared a Public Health Emergency of International Concern, new communication channels have also been built between Ministries of Health from different countries14. As the world becomes more interconnected and urbanized, it is likely that many other epidemics will follow. Therefore, it is paramount that lessons learned from Zika lead to fast and effective responses to future global threats.",
"appendix": "Author contributions\n\n\n\nTazio Vanni contributed to conception, drafting, and submission of the manuscript.\n\nKarlos Diogo Chalegre contributed to the conception and drafting of the manuscript.\n\nCamile Giaretta Sachetti contributed to the conception and drafting of the manuscript.\n\nPedro Reginaldo dos Santos Prata contributed to the conception and drafting of the manuscript.\n\nMarco Antônio de Araújo Fireman contributed to the conception and drafting of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMinistério da Saúde, Brazil: Boletim Epidemiológico – Monitoramento dos casos de dengue, febre do Chikungunya e febre pelo vírus Zika até a Semana Epidemiológica 20 de 2016. 2016. (accessed Jul 14, 2016). Reference Source\n\nMinistério da Saúde, Brazil: Informe Epidemiológico Nº 11 – Semana Epidemiológica (SE) 27/2016 (03 A 09/07/2016) Monitoramento Dos Casos De Microcefalia No Brasil. 2016. (accessed Jul 14, 2016). Reference Source\n\nMinistério da Saúde, Brazil: Portaria GM nº 1.813, de 11 de Novembro de 2015. 2015. (accessed Jul 14, 2016). Reference Source\n\nOliveira Melo AS, Malinger G, Ximenes R, et al.: Zika virus intrauterine infection causes fetal brain abnormality and microcephaly: tip of the iceberg? Ultrasound Obstet Gynecol. 2016; 47(1): 6–7. PubMed Abstract | Publisher Full Text\n\nGarcez PP, Loiola EC, Madeiro da Costa R, et al.: Zika virus impairs growth in human neurospheres and brain organoids. Science. 2016; 352(6287): 816–8. PubMed Abstract | Publisher Full Text\n\nBrasil P, Pereira JP Jr, Raja Gabaglia C, et al.: Zika Virus Infection in Pregnant Women in Rio de Janeiro - Preliminary Report. N Engl J Med. 2016. PubMed Abstract | Publisher Full Text\n\nFrança GV, Schuler-Faccini L, Oliveira WK, et al.: Congenital Zika virus syndrome in Brazil: a case series of the first 1501 livebirths with complete investigation. Lancet. 2016; pii: S0140-6736(16)30902-3. PubMed Abstract | Publisher Full Text\n\nDutra HL, Dos Santos LM, Caragata EP, et al.: From lab to field: the influence of urban landscapes on the invasive potential of Wolbachia in Brazilian Aedes aegypti mosquitoes. PLoS Negl Trop Dis. 2015; 9(4): e0003689. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbad-Franch F, Zamora-Perea E, Ferraz G, et al.: Mosquito-disseminated pyriproxyfen yields high breeding-site coverage and boosts juvenile mosquito mortality at the neighborhood scale. PLoS Negl Trop Dis. 2015; 9(4): e0003702. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAyres CF: Identification of Zika virus vectors and implications for control. Lancet Infect Dis. 2016; 16(3): 278–279. PubMed Abstract | Publisher Full Text\n\nWHO: Target Product Profiles for better diagnostic tests for Zika virus infection. 2016. (accessed Jul 14, 2016). Reference Source\n\nPhase I, Open-label, Dose-Ranging Study to Evaluate the Safety, Tolerability, and Immunogenicity of GLS-5700 Administered ID Followed by EP in Dengue Virus-Naïve Adults. ClinicalTrials.gov processed this record on July 13, 2016. (accessed Jul 14, 2016). Reference Source\n\nMinistério da Saúde, Brazil: Portaria GM nº 1.046, de 20 de Maio de 2016. 2016. (accessed Jul 14, 2016). Reference Source\n\nWHO: WHO statement on the first meeting of the International Health Regulations (2005) (IHR 2005) Emergency Committee on Zika virus and observed increase in neurological disorders and neonatal malformations. 2016. (accessed Jul 14, 2016). Reference Source"
}
|
[
{
"id": "15505",
"date": "08 Aug 2016",
"name": "Luciano Pamplona de Góes Cavalcanti",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nBelow my considerations. I believe that the article can be indexed with minor adjustments.\nI do not agree with the statement: \"In the last decades, Brazil has faced different arbovirus epidemics. However, none of them had the complexity of Zika virus and associated diseases\". Dengue has caused and still causes serious public health and has much higher mortality rate in Brazil.\nThe article cites many epidemiological bulletin of the Ministry of Health. This is important, but it has some articles that deserve to be mentioned. As an example, we have published papers citing cases of zika in Brazil (RN and BA) before the publication of the Ministry of Health report.\nEven with all the advances cited by the authors I believe that it is necessary to mention the delay in the release of funds for research on zika in Brazil.",
"responses": []
},
{
"id": "15506",
"date": "17 Aug 2016",
"name": "Luciano A. Moreira",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comment\n\nVanni et al. describe in this manuscript what has been, the Brazilian Government response, since the declaration of the National Public Health Emergency due to Zika. It is well written, concise and indicates the Brazilian Government is aware and is trying to tackle the Zika epidemics by using a multidisciplinary approach. It is important to guarantee though continuous funding for these different strategies.\n\nSpecific comment\n\nSecond paragraph: It is not clear whether the research agenda falls into the third pillar described by the authors on paragraph 1 3) Research & education. If yes, where education is inserted in the four main areas?",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1915
|
https://f1000research.com/articles/5-125/v1
|
01 Feb 16
|
{
"type": "Case Report",
"title": "Case Report: Frontalis sign for early bedside diagnosis of impending uncal herniation",
"authors": [
"Sunil Munakomi",
"Bijoy Mohan Kumar",
"Bijoy Mohan Kumar"
],
"abstract": "It is prudent to have early diagnosis and timely management of uncal herniation for better management of neurosurgical patients. There are several clinical and radiological armamentarium that aid in early recognition of the condition. Through this case report, we try to highlight a simple bedside clinical sign that can help in early recognition of the impending uncal herniation. The improvement in the sign also confirms the resolution of the mass effect in the postoperative period. This is especially helpful for doctors working in the periphery or in resource restrained areas, for a timely referral of the patient to tertiary centre.",
"keywords": [
"Trauma",
"herniation",
"sign"
],
"content": "Introduction\n\nTraumatic brain injury (TBI) is now a global epidemic1. The prognosis of patients with head injury is dependent on many clinical parameters but one of the major determinants is the time lapsed for appropriate management2. TBI has a significant impact not only on the patients and their relatives but also has a major influence on the health and socioeconomical status in the global arena. Thus, it is prudent to have clinical tools for early recognition of life threatening neurosurgical emergencies. Herein we discuss one such example: Frontalis sign for detecting early uncal herniation. This may be helpful for early referral of patients to tertiary care centres and for timely management of the same. It could therefore have a positive impact on these patients, resulting in a better outcome.\n\n\nCase report\n\nA 50-year-old male from Siraha, a distant village in Nepal, was referred to our neurosurgical centre following a road traffic accident after being hit by a speeding car. The patient had a brief loss of consciousness and a single episode of vomiting following the incident. There was no history of seizurogenic activity observed during the transfer. On arrival to the emergency department, his Glasgow coma scale (GCS) was E3M6V5 with no paucity in movement of any limbs. His vital parameters were within normal range with blood pressure of 130/90, pulse rate of 86/min and oxygen saturation of 99% in room air. It was difficult to assess differences in pupillary size as he had corneal opacity on the left eye, resulting from an injury sustained during his childhood. However, on close examination, we observed that there was prominence of the forehead wrinkles on the right half of his face especially when the patient was trying to open his eyes during conversation, which we termed as frontalis sign (Figure 1). The wrinkles on the contra lateral half were normal with no abnormal deviation of angle of the mouth dismissing the differential diagnosis of upper facial nerve palsy. Because of the finding, we suspected impending uncal herniation in the patient and thereby advised for an emergency computed tomography (CT) scan of the head. It revealed right sided huge temporo-parietal contusion with thin fronto-temporo-parietal subdural hematoma with features of uncal herniation (Figure 2). The condition was explained to his relatives and they were counseled for emergency evacuation of the hematoma. On their consent, we performed a craniotomy, evacuation of the subdural hematoma and removal of the contusion. Following the procedure, the brain was lax and pulsatile. The patient was extubated without any untoward events in the postoperative period. The frontalis sign diminished following the surgery (Figure 3). The post operative scan confirmed resolution of the mass effect and normalization of the cisternal anatomy (Figure 4). The patient was started on Levtiracetam 500 mg intravenously every 12 hours which was changed to oral medication after three days as seizure prophylaxis. The patient was discharged after suture removal on the 8th postoperative day. The patient followed up in the outpatient clinic 2 weeks later in sound health. Eye opening was near normal. The patient was advised for monthly follow up.\n\n\nDiscussion\n\nLevator palpebrae superioris supplied by the third nerve helps in elevation of the lid during eye opening4,5. However third nerve involvement due to uncal herniation weakens the muscle thereby restricting its action6. In order to compensate the deficits, the frontalis belly of the occipito-frontalis muscle supplied by the facial nerve, helps in elevation of the lid7,8. This leads to prominence of forehead wrinkles in the same side on comparison to the other half. This is termed frontalis sign. This can act as a reliable bedside marker for diagnosing impending herniation. There are other routine signs for impending uncal herniation such as anisocoria. But there may be inter-observer bias in assessing the same. Sometimes drugs such as Ipratropium Bromide used for nebulisation in the intensive care unit can cause anisocoria. It is also difficult to observe the size and reaction of the pupils in patients with severe eye lid swellings9. The frontalis sign can be used as an adjunct for diagnosing uncal herniation and thereby initiating the correct management. This is even more valuable for proper patient referral from peripheral and resource limited setups, especially in developing countries like ours who are still far behind implementing the guidelines for managing patients with TBI10.\n\n\nConclusion\n\nThe implications of the use of this simple bedside sign for early diagnosis of the uncal diagnosis can be influential in providing timely and correct therapeutic targets for patients with TBI. It can be a valuable adjunct to the present panoply of our armamentarium in diagnosis the traumatic cerebral herniation syndromes.\n\n\nConsent\n\nWritten informed consent was obtained from the daughter of the patient for publication of this case report and any accompanying images and/or other details that could potentially reveal the patient’s identity.",
"appendix": "Author contributions\n\n\n\nSM reviewed the literature and formatted the paper. BB suggested, revised and edited the final format.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nJennett B: Epidemiology of head injury. J Neurol Neurosurg Psychiatry. 1996; 60(4): 362–369. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJennett B, Teasdale G, Braakman R, et al.: Prognosis of patients with severe head injury. Neurosurgery. 1979; 4(4): 283–289. PubMed Abstract\n\nHumphreys I, Wood RL, Phillips CJ, et al.: The costs of traumatic brain injury: a literature review. Clinicoecon Outcomes Res. 2013; 5: 281–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEttl A, Priglinger S, Kramer J, et al.: Functional anatomy of the levator palpebrae superioris muscle and its connective tissue system. Br J Ophthalmol. 1996; 80(8): 702–707. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNg SK, Chan W, Marcet MM, et al.: Levator palpebrae superioris: an anatomical update. Orbit. 2013; 32(1): 76–84. PubMed Abstract | Publisher Full Text\n\nMaramattom BV, Wijdicks EF: Uncal herniation. Arch Neurol. 2005; 62(12): 1932–1935. PubMed Abstract | Publisher Full Text\n\nBérzin F: Occipitofrontalis muscle: functional analysis revealed by electromyography. Electromyogr Clin Neurophysiol. 1989; 29(6): 355–8. PubMed Abstract\n\nKushima H, Matsuo K, Yuzuriha S, et al.: The occipitofrontalis muscle is composed of two physiologically and anatomically different muscles separately affecting the positions of the eyebrow and hairline. Br J Plast Surg. 2005; 58(5): 681–7. PubMed Abstract | Publisher Full Text\n\nYalcin S, Pampal K, Erden A, et al.: Do we really need to panic in all anisocoria cases in critical care? Indian J Anaesth. 2010; 54(4): 365–366. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNational Collaborating Centre for Acute Care (UK): Head Injury: Triage, Assessment, Investigation and Early Management of Head Injury in Infants, Children and Adults. London: National Collaborating Centre for Acute Care (UK), (NICE Clinical Guidelines, No. 56.), 2007. PubMed Abstract"
}
|
[
{
"id": "13921",
"date": "20 May 2016",
"name": "G. Bryan Young",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper is of interest. The authors should give details of pupillary reactivity. With herniation there is usually a change in pupillary size and reactivity before the ptosis. Was this not the case here?",
"responses": [
{
"c_id": "2053",
"date": "04 Jul 2016",
"name": "Sunil Munakomi",
"role": "Author Response",
"response": "We thank you for the report on our article.We acknowledge the role of assessing pupillary size in patients with trauma. Here, we are emphasizing on the adjunct role of frontalis sign in diagnosing early uncal herniation in cases wherein pupillary assessment is problematic as in severe eye lid swelling,corneal injuries (in our case), following use of sedatives or following post traumatic seizure."
}
]
},
{
"id": "15318",
"date": "29 Jul 2016",
"name": "Osman Sinanović",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this case report authors try to highlight a simple bedside clinical sign (frontalis sign) in early recognition of the impending uncul herniation.\n\nTitle “Frontalis sign for early bedside diagnosis of impending uncal herniation\" is appropriate for content of the article but with a clinical sign we can not mkae adiagnosis. So, title “Frontalis sign for early bedside consideration of impending uncal herniation” may be more appropriate.\n\nArticle content including design, case report and discussion is correct.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-125
|
https://f1000research.com/articles/5-1914/v1
|
04 Aug 16
|
{
"type": "Opinion Article",
"title": "Navigating the Zika panic",
"authors": [
"Nathan D. Grubaugh",
"Kristian G. Andersen"
],
"abstract": "The epidemics of Ebola virus in West Africa and Zika virus in America highlight how viruses can explosively emerge into new territories. These epidemics also exposed how unprepared we are to handle infectious disease emergencies. This is also true when we consider hypothesized new clinical features of infection, such as the associations between Zika virus infection and severe neurological disease, including microcephaly and Guillain-Barré syndrome. On the surface, these pathologies appear to be new features of Zika virus infection, however, causal relationships have not yet been established. Decades of limited Zika virus research are making us scramble to determine the true drivers behind the epidemic, often at the expense of over-speculation without credible evidence. Here we review the literature and find no conclusive evidence at this time for significant biological differences between the American Zika virus strains and those circulating elsewhere. Rather, the epidemic scale in the Americas may be facilitated by an abnormally warm climate, dense human and mosquito populations, and previous exposure to other viruses. Severe disease associated with Zika virus may therefore not be a new trait for the virus, rather it may have been overlooked due to previously small outbreaks. Much of the recent panic regarding Zika virus has been about the Olympics in Brazil. We do not find any substantial evidence that the Olympics will result in a significant number of new Zika virus infections (~10 predicted) or that the Olympics will promote further epidemic spread over what is already expected. The Zika virus epidemic in the Americas is a serious situation and decisions based on solid scientific evidence - not hyped media speculations - are required for effective outbreak response.",
"keywords": [
"Zika virus",
"arbovirus",
"virus emergence",
"microcephaly",
"antibody-dependent enhancement",
"Brazil Olympics 2016"
],
"content": "Zika’s path from obscurity to severity\n\nZika virus, a name now synonymous with birth defects by many people, was not always a topic of public health debate. In fact, for 67 years, the virus was mostly ignored (89 publications from 1947 to 2014, compared to 850 over the last 19 months). That is because when Zika virus was discovered in 1947 it was not thought to cause severe enough disease in humans to warrant intense research1. Fast forward to today and people are talking about canceling one of the world’s most watched events, the Olympics, due to the Zika virus epidemic in Brazil2,3. So, what happened? Did the virus change? Did we misinterpret its threat from the beginning? And will the Olympics this summer really exacerbate the current epidemic or provide new opportunities for Zika virus emergence? Zika virus research is now pouring in fast, but at times at the expense of fast-tracked studies and misinterpretation of results. As a result, there is significant confusion surrounding the Zika virus epidemic and many of the core questions need to be revisited.\n\nIt has been suggested that “The Brazilian strain of Zika virus harms health in ways that science has not observed before”3 and “[Africa and Asia] have mostly avoided the post-2013 neurotrophic strains of the virus that are ravaging Brazil”2. Based on available evidence, however, it is too early to say whether this strain is in fact fundamentally different from other Zika virus strains. Only recently has Zika virus been associated with large outbreaks (since 2007 - Yap Island4) and severe disease such as microcephaly and Guillain-Barré syndrome (since 2013 - French Polynesia5). The epidemic in the Americas has proven to alarmingly increase these trends - 0.5 to 1.5 million suspected infections and ~4,000 cases of microcephaly in Brazil alone6. What are the real reasons behind the severity of this epidemic? We will explore aspects of 1) viral genetics that might alter transmission and pathogenicity in humans, 2) the ecological conditions in the Americas, 3) the potential impact of dengue virus on Zika virus-associated pathology, and 4) how small sample sizes and under reporting may have skewed our previous assumptions of Zika virus and the disease it can cause. Using this knowledge, we will discuss whether a global event like the Olympics would really impact further Zika virus emergence and the expansion of the epidemic.\n\n\nIs Zika virus different today than it was when it was first discovered?\n\nUndoubtedly, yes, Zika virus circulating today is genetically different from the Zika virus of the past. A key aspect of Zika virus is that it has an RNA genome. Central features of RNA virus biology is that these viruses replicate, produce large population sizes, but do so with lots of errors (mostly because their polymerases lack proofreading mechanisms, adding ~1 mutation per genome replication)7–9. Therefore, all RNA viruses have the ability to evolve fast relative to most DNA-based organisms10, and Zika virus has evolved into at least two distinct lineages: African and Asian11. The viruses circulating in the Americas belong to the Asian lineage, which, to the best of our knowledge, originated in East Africa12. Comparing the genetics of the first discovered Zika virus strain from 1947 (Uganda, strain MR766) to the strain currently circulating in the Americas (2015 Puerto Rico, strain PRVABC-59) reveals mutations in >1,100 nucleotide positions (~89% similarity), and confirms that yes, the viral genome is different. While the genetic makeup has changed - as is expected - the more important question, however, is whether this means that the currently circulating strain of Zika virus has a fundamentally different “behavior” (i.e., phenotype)?\n\nUnfortunately, we are critically lacking comparative studies to directly address whether genetic changes in the virus are significantly contributing to the epidemic. For example, are there differences in mosquito vector competence (i.e., ability of a population of mosquitoes to transmit the virus) between Zika virus strains in Africa and the Americas? While studies have shown that competence of the suspected Zika virus vectors in the Americas, Aedes aegypti and Ae. albopictus, can vary between mosquito populations13, the influence of Zika virus genetics has not been tested. There is, however, precedence for mosquito-borne viruses to adapt to local mosquitoes, increasing their epidemic potential. This happened with chikungunya virus during the Indian Ocean epidemic14 and West Nile virus during its invasion of the United States15,16. Hence there could be some yet-to-be discovered Zika virus mutations that facilitate faster transmission and increased rates of mosquito infection. A lot of work needs to be directed towards lab and field mosquitoes studies to actually determine if this has occurred and whether Zika virus could have adapted to enhance local transmission. At this point, however, there is no evidence that the Zika viruses in the Americas have adapted to the local conditions or can be transmitted any more efficiently than previous strains.\n\nPerhaps even more controversial and urgent are the questions: is the Brazilian strain of Zika virus more capable of 1) being transmitted during pregnancy or 2) causing neuropathogenesis leading to complications such as microcephaly17 and Guillain-Barré syndrome5 than strains from Africa or other Asian strains? Several recent laboratory studies have shown that Zika virus can infect placental cells18–20, be vertically transmitted to offspring during pregnancy21,22, target and replicate in neuronal cells20,23,24, and cause birth defects25 (in vivo in mice, in vitro with human cells, organoids, and organ explants). These studies, however, were conducted with a variety of Zika virus strains indicating that some of these phenotypes are common features of the virus, irrespective of the strain. In fact, the virus was first isolated in 1947 by injecting serum from a febrile rhesus macaque directly into a mouse brain1 and a subsequent paper published in 1971 showed that the same Zika virus strain could cause neurological disease in mice26. While these observations may not be that surprising - the mice were infected intracerebrally after all (as is common for these types of experiments) - these early experiments already demonstrated that Zika virus can replicate and cause pathology in neurons. Together, these studies suggest that vertical transmission and neuropathogenesis are not specific attributes of the Brazilian strain and perhaps Zika virus has always been capable of this.\n\nSo were those ancestral strains from the 1940’s to 1970’s in Africa reported to cause mild disease actually misunderstood? We know that people in some areas of Africa had high seroprevalence to Zika virus (e.g., ~30% in Nigeria during the early 1970s27). Perhaps disease associated with Zika virus was just overlooked in Africa because of the many other diseases such as malaria, acquired immune deficiency syndrome (AIDS), and tuberculosis that ravage the continent. In the Americas, a large number of Zika virus-naive people (i.e., without previous immunity) are getting infected at the same time, which may reveal previously unknown clinical features of viral infection. Finally, genetic analysis of Zika virus strains has yet to discover any appreciable patterns associated with adaptation towards humans, vector species, or disease outcome28. This is not to say that it has not occurred, only that at this point in time our sampling is too insufficient to make any conclusions. Therefore, more experimental evidence is required before we can say whether Zika virus genetics or phenotype has changed in any significant way.\n\n\nWhy is the epidemic in the Americas so bad?\n\nZika virus is not the first, nor likely the last mosquito-borne virus, to explosively emerge in the Americas. In 1999, West Nile virus was introduced into the New York area and quickly spread across the continent, killing thousands of people and millions of birds (reviewed by 29). Even more recently, in 2013, chikungunya virus emerged throughout the Caribbean and much of the tropical regions in the Americas (reviewed by 30). By 2015, there were already more than one million suspected cases31. Since chikungunya and Zika viruses share similar ecologies (humans and Ae. aegypti), the current Zika virus outbreak should not be so surprising, given recent histories. Even the 2007 Yap Island outbreak gave us some indication of its potential - it is estimated that 73% of the population became infected with Zika virus4. A large outbreak in the Americas almost seemed inevitable, but why?\n\nWell, likely because the Americas are home to large and dense populations of hosts (humans) without previous Zika virus immunity, and vectors (mosquitoes) capable of transmission. The climate may also have contributed to the scale and intensity of the epidemic; 2015 was the warmest year on record in the Americas32, which could have enhanced Zika virus transmission. Warmer temperatures can increase mosquito abundance, survival, blood feeding rates, and vector competence33–36. Therefore, the extreme circumstances caused by El Niño and global climate change may have contributed to a higher density of mosquitoes37. Together, these factors represent an ideal recipe for an infectious disease epidemic.\n\nThe unfortunate surprise was the discovery of an association between severe neurological complications and Zika virus infection, especially among newborns38. This, however, could just be a consequence of numbers and reporting. Previous outbreaks may have missed these links because they were too small. In Brazil, the current estimate is that between 1–13% of pregnant women who become infected with Zika virus in their first trimester will deliver babies with microcephaly17,39. That is 8-650× the baseline microcephaly rate of 0.02–0.12% per live birth17,39. The largest previous Zika outbreak - which occurred in French Polynesia - had an estimated size of 30,000 infections based on serological evidence (11% of the 270,000 people)40. Retrospective analysis of first trimester pregnancies indicated that about 1% of Zika virus infections resulted in babies born with microcephaly - a total of eight cases41. During the Yap Island outbreak, researchers estimated that about 5,000 of the 7,000 inhabitants over the age of three became infected4. Based on Yap State census reports42, roughly 200–300 women may have been pregnant during the outbreak, and only about ⅓ would have been in the first trimester. If the previous estimates were accurate during the Yap Island outbreak, then it may be possible that no babies were born with microcephaly just because there were not enough infected pregnant women for the chance occurrence. The current epidemic in the Americas, including Brazil, may therefore just seem more severe because there are more infected people to detect rare pathological complications such as microcephaly.\n\nDiscovering new disease associated with a particular virus only after a large outbreak is not unique to Zika virus. In fact, we can learn from previous epidemics with different viruses, such as West Nile virus. Prior to the 1990’s, West Nile virus was only known to cause sporadic outbreaks with a few cases of severe neurological disease. An outbreak in Romania43, however, redefined our perception of the virus. From 1996–1997, there were more than 500 clinical West Nile cases with a fatality rate approaching 10%, representing the largest mosquito-borne virus outbreak in Europe in more than a decade. Other outbreaks in urban areas soon followed, occurring in Russia44, Israel45, and the United States46, all of which included neurological disease in about 1% of the cases29. While genetic changes to the virus may account for some of the increase in severity47, most of it can be attributed to previously underestimating its severity due to small sample sizes. Many parallels can be made between what happened with West Nile virus and the sudden increase in Zika virus associated disease during its current emergence.\n\nThere may also be immunological explanations for pathology associated with Zika virus infection in the Americas. Zika and dengue viruses co-occur in many parts of the world. The fastest growing numbers of dengue cases occur in Latin America and the Caribbean with more than 10 million apparent infections a year48, a ~250% increase since 199049. One interesting hypothesis is that antibodies produced from a previous dengue virus infection may enhance subsequent Zika virus infection50–53. The proposed mechanism is that antibodies targeting dengue virus can bind to Zika virus during an active infection, but cannot always neutralize it. Instead, the bound antibodies can actually help Zika virus infect monocytes by attaching to the cell surface receptors (Fc gamma) and mediating efficient entry. This process of antibody-dependent enhancement is also known to occur between different serotypes of dengue virus and is a risk factor for severe dengue disease (reviewed by 54). Since 2010, between 600,000 and 1.6 million annual dengue virus cases in Brazil have been reported55. Therefore the high incidence of dengue virus infection may be increasing the observed pathogenicity of Zika virus in the Americas. On the other hand, dengue and Zika viruses co-occur elsewhere, so the Americas may not be so unique. Indeed, further research is urgently needed to determine if dengue virus is not only exacerbating the Zika virus epidemic in the Americas, but also anywhere the two viruses co-circulate.\n\n\nHow many visitors will become infected with Zika virus during the Olympics?\n\nNow turning our attention from the biology and genetics of Zika virus, to the different risks associated with Zika virus and the Olympics. There are two main risks to consider: 1) the risk of further spread and 2) personal risk to visitors. These are two very different questions, but often they get blurred together. Here we will discuss them separately. First, how many of the expected half a million visitors to Rio de Janeiro during the summer Olympics this August will become infected with Zika virus? Massad et al. predict that the numbers of individuals acquiring Zika virus during the Games is low - 1 in ~30,000 to 100,000 people56. This translates to only 5 to 15 visitors during the 3-week games will locally acquire Zika virus. The same model was used to predict that 3 to 59 visitors would become infected with dengue virus during the 2014 World Cup57. The actual reported number was three, suggesting that such estimates are relatively accurate58. Another group estimated that 3 to 80 visitors will become infected with Zika virus during the Games59. Based on previous experience and scientific evidence, we might therefore only expect that about 10 people traveling to Brazil for the Games will get infected with Zika virus. Compared to the overall number of cases, that is an astonishingly low number.\n\nOne main reason behind the few predicted Zika infections of visitors is that August is winter in Brazil, so mosquito densities will have declined60,61. That alone will severely decrease the likelihood of exposure to Zika virus through fewer mosquito “bites”. The remaining risk is dependent upon the 1) mosquito species, 2) proportion of infected mosquitoes, and 3) transmission rates upon blood feeding. A recent report helped to validate our assumption that Ae. aegypti is vectoring Zika virus in at least some parts of the Americas62, though other species may be involved63. The proportion of Ae. aegypti that are infected with Zika virus at any given time, however, is unclear. Ae. aegypti infection rates range from extremely low (unpublished data suggesting only a few Zika virus RNA-positive mosquitoes among thousands tested) to very high (5–17% near homes of suspected Zika patients62). If the infection rates are similar to chikungunya or dengue viruses, then we can expect that 1–2% of Ae. aegypti are infected with Zika virus64,65. Moreover, only a small portion (5–50%) of infected mosquitoes can actually transmit the virus13. So even if you get fed upon by hundreds of mosquitoes, odds are that you will not get exposed to Zika virus.\n\n\nWill the Olympics enhance the further spread of Zika virus?\n\nThe world is interconnected. Zika virus and many other mosquito-borne viruses have already utilized this interconnectivity to travel great distances. Does a global event like the Olympics really enhance this problem? One estimate indicates that 100 to 400 people infected with Zika virus will enter Europe in 2016 due to normal travel from endemic regions66. That is already 7–80× greater than the number of people predicted to become infected during the Games (~10 - see above). In reality, not enough people will get infected with Zika virus while visiting Brazil for three weeks to have a significant impact on the already expected viral spread. Unfortunately, Zika virus will spread, just as dengue, chikungunya, and West Nile viruses have done before. Holding the Olympic Games in Brazil will have no, or extremely limited, effect on this process.\n\nTo really understand the risks, let’s use an example. If a person is infected with Zika virus and returns to their home country, what are the real chances that the infected person could initiate local mosquito-borne transmission? The answer is largely dependent on the local conditions. Specifically, does the environment support enough competent mosquitoes that feed on humans to facilitate transmission? And will such transmission be sustained? Most of Europe, the United States, and other temperate regions cannot support local Zika virus transmission because they either 1) have an environment that is inhospitable to Ae. aegypti (or other susceptible mosquitoes) or 2) have infrastructure to minimize the risks of mosquito exposure (e.g., air-conditioned homes and mosquito management programs)67. Much of the tropical and subtropical regions of the world, however, have suitable environmental conditions to support Zika virus transmission67. Spread to many of these places is not concerning, because they already have autochthonous (local) transmission of Zika virus. The Centers for Disease Control and prevention (CDC) recently conducted an assessment of countries that could be at risk of Zika virus importation following the Olympics68. They suggest that Angola and China (among other countries) could be at risk because 1) they are currently not reporting autochthonous Zika virus transmission, 2) they likely have conditions to support transmission (i.e., dense human populations and Ae. aegypti), and 3) there is a high amount of expected air travel from Brazil. In short, it really depends on the home country of the traveler, what season it is, their economic status, whether they can protect themselves from mosquitoes, and many other variables. While a single traveler could be responsible for a new outbreak (as suggested for the Zika virus introduction into Brazil28), in all likelihood these events are extremely rare.\n\nThe ‘single introduction’ hypothesis put forward by Faria et al.28 has often (wrongly) been used to suggest that it only takes one infected traveler to start an outbreak (i.e., giving the sense that this could happen anytime)2. That is not correct and was also not suggested by the authors. Instead, what the Faria et al. data show, is that the chance of starting an outbreak is extremely low. If it had been high, we would have seen multiple introductions of Zika virus into Brazil (and elsewhere), due to travelers arriving from Zika endemic countries. We don’t see that, hence it likely takes many - not just one - infected travelers for the chance occurrence to start an outbreak.\n\nDirect human-to-human transmission is another possible route of Zika virus infection. These routes notably include transmission from mother to child during pregnancy and sexual transmission from a man to a woman69. Other forms of human-to-human transmission scenarios also appear to exist70. Therefore, could sustained Zika virus transmission occur without mosquitoes and should this be a concern for further spread of the epidemic? Again, let’s use an example: an infected man returns home from the Games and has sex with his partner. There are numerous reports of Zika virus infection associated with sex with a man (or woman71) returning from an endemic region72–74. Therefore, in this scenario, there is immediate risk to his partner. Importantly, however, in each of these reports, Zika virus spread was limited to just those single contacts. Thus, sex and other modes of direct contact with an infectious individual is highly unlikely to lead to sustained transmission in a new population. It has also been estimated that the role of sexual transmission in Brazil is minimal compared to mosquitoes75, and without mosquitoes, transmission would dissipate. The single most compelling piece of evidence to support that Zika virus is primarily mosquito-borne is that it is only known to occur in regions with Ae. aegypti67. Therefore, Zika virus is still considered to be primarily transmitted by mosquitoes and sexual transmission (or other yet-to-be-discovered human-to-human means of transmission) will likely not expand the expected range of the Zika virus epidemic.\n\n\nTake-home message\n\nOur rapidly expanding knowledge about Zika virus is starting to reveal important information about the current epidemic and suggests that we may have misjudged its epidemic potential for decades. We explored four key areas to demonstrate how the epidemic severity may be more related to the conditions in the Americas rather than new disease caused by Zika virus. 1) There is currently no definitive evidence that the strain of Zika virus in Brazil has altered potential for transmission or pathogenicity in humans compared to the strains circulating in Africa and Asia (although this does not mean that the Brazilian strain does not have an altered phenotype compared to other strains, only that no good evidence is currently available to suggest that is the case). 2) Major factors for the scope of the epidemic were likely large urban settings housing people without immunity and an abnormally warm climate leading to a large population of mosquitoes. 3) Previous exposure to dengue virus could increase Zika virus disease severity, though such a connection is yet to be demonstrated as an important risk factor. 4) The recent associations of some Zika virus infections with severe neurological conditions, such as microcephaly and Guillain-Barré syndrome, may be simply a reflection of sample sizes - large numbers of infections are often required to discover rare pathologies.\n\nThe risks regarding Zika virus and the Olympic Games in Brazil are 1) whether it will enhance the epidemic spread and 2) personally to people attending the games. The numbers of Olympic visitors expected to get infected with Zika virus in Brazil and travel home is far lower than the total numbers of these occurrences already expected to happen throughout the year. Therefore, the Olympics will not be a significant conduit for further epidemic spread. There is a personal risk of infection, though it is also predicted to be low. Obviously, pregnant women have the greatest risk as they could pass the virus to their developing fetus, with the possibility of causing severe neurological complications. Therefore each family needs to evaluate the consequences and likelihood of Zika virus infection to determine if they should travel to any region of the world with active Zika virus transmission. The Zika virus epidemic is a severe problem, but decisions should be based on scientific evidence78,79 and not fear-mongering2. These should be lessons to keep in mind when we argue about some other relatively unknown virus before the start of Tokyo Olympics in 2020.",
"appendix": "Author contributions\n\n\n\nNDG and KGA conceived the ideas and wrote the manuscript.\n\n\nCompeting interests\n\n\n\nAll authors declare that they have no competing interests.\n\n\nGrant information\n\nNDG is supported by a National Institutes of Health (NIH) training grant 5T32AI007244-33. KGA is a PEW Biomedical Scholar, and his work is supported by an NIH National Center for Advancing Translational Studies Clinical and Translational Science Award UL1TR001114, and National Institute of Allergy and Infectious Diseases (NIAID) contract HHSN272201400048C.\n\n\nReferences\n\nDick GW, Kitchen SF, Haddow AJ, et al.: Zika virus. I. Isolations and serological specificity. Trans R Soc Trop Med Hyg. Elsevier; 1952; 46(5): 509–520. PubMed Abstract | Publisher Full Text\n\nAttaran A: Zika virus and the 2016 Olympic Games. Lancet Infect Dis. Elsevier; 2016; pii: S1473-3099(16)30230-4. PubMed Abstract | Publisher Full Text\n\nRio Olympics Later [Internet]. [cited 25 Jul 2016]. Reference Source\n\nDuffy MR, Chen TH, Hancock WT, et al.: Zika virus outbreak on Yap Island, Federated States of Micronesia. N Engl J Med. 2009; 360(24): 2536–2543. PubMed Abstract | Publisher Full Text\n\nCao-Lormeau VM, Blake A, Mons S, et al.: Guillain-Barré Syndrome outbreak associated with Zika virus infection in French Polynesia: a case-control study. Lancet. 2016; 387(10027): 1531–1539. PubMed Abstract | Publisher Full Text\n\nECDC: Rapid risk assessment: Zika virus disease epidemic, seventh update [Internet]. [cited 25 Jul 2016]. Reference Source\n\nHolland J, Spindler K, Horodyski F, et al.: Rapid evolution of RNA genomes. Science. 1982; 215(4540): 1577–1585. PubMed Abstract | Publisher Full Text\n\nSteinhauer DA, Domingo E, Holland JJ: Lack of evidence for proofreading mechanisms associated with an RNA virus polymerase. Gene. 1992; 122(2): 281–288. PubMed Abstract | Publisher Full Text\n\nDomingo E, Sabo D, Taniguchi T, et al.: Nucleotide sequence heterogeneity of an RNA phage population. Cell. Elsevier; 1978; 13(4): 735–744. PubMed Abstract | Publisher Full Text\n\nDuffy S, Shackelton LA, Holmes EC: Rates of evolutionary change in viruses: patterns and determinants. Nat Rev Genet. 2008; 9(4): 267–276. PubMed Abstract | Publisher Full Text\n\nLanciotti RS, Lambert AJ, Holodniy M, et al.: Phylogeny of Zika Virus in Western Hemisphere, 2015. Emerg Infect Dis. 2016; 22(5): 933–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaddow AD, Schuh AJ, Yasuda CY, et al.: Genetic characterization of Zika virus strains: geographic expansion of the Asian lineage. PLoS Negl Trop Dis. 2012; 6(2): e1477. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChouin-Carneiro T, Vega-Rua A, Vazeille M, et al.: Differential Susceptibilities of Aedes aegypti and Aedes albopictus from the Americas to Zika Virus. PLoS Negl Trop Dis. 2016; 10(3): e0004543. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsetsarkin KA, Vanlandingham DL, McGee CE, et al.: A single mutation in chikungunya virus affects vector specificity and epidemic potential. PLoS Pathog. 2007; 3(12): e201. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoudy RM, Meola MA, Morin LL, et al.: A newly emergent genotype of West Nile virus is transmitted earlier and more efficiently by Culex mosquitoes. Am J Trop Med Hyg. 2007; 77(2): 365–370. PubMed Abstract\n\nEbel GD, Carricaburu J, Young D, et al.: Genetic and phenotypic variation of West Nile virus in New York, 2000–2003. Am J Trop Med Hyg. ASTMH; 2004; 71(4): 493–500. PubMed Abstract\n\nJohansson MA, Mier-y-Teran-Romero L, Reefhuis J, et al.: Zika and the Risk of Microcephaly. N Engl J Med. 2016; 375(1): 1–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTabata T, Petitt M, Puerta-Guardo H, et al.: Zika Virus Targets Different Primary Human Placental Cells, Suggesting Two Routes for Vertical Transmission. Cell Host Microbe. 2016; pii: S1931-3128(16)30300-6. PubMed Abstract | Publisher Full Text\n\nQuicke KM, Bowen JR, Johnson EL, et al.: Zika Virus Infects Human Placental Macrophages. Cell Host Microbe. 2016; 20(1): 83–90. PubMed Abstract | Publisher Full Text\n\nRetallack H, Di Lullo E, Arias C, et al.: Zika Virus in the Human Placenta and Developing Brain: Cell Tropism and Drug Inhibition [Internet]. bioRxiv. 2016; 058883. Publisher Full Text\n\nWu KY, Zuo GL, Li XF, et al.: Vertical transmission of Zika virus targeting the radial glial cells affects cortex development of offspring mice. Cell Res. 2016; 26(6): 645–654. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiner JJ, Cao B, Govero J, et al.: Zika Virus Infection during Pregnancy in Mice Causes Placental Damage and Fetal Demise. Cell. 2016; 165(5): 1081–1091. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarcez PP, Loiola EC, Madeiro da Costa R, et al.: Zika virus impairs growth in human neurospheres and brain organoids. Science. 2016; 352(6287): 816–818. PubMed Abstract | Publisher Full Text\n\nHanners NW, Eitson JL, Usui N, et al.: Western Zika Virus in Human Fetal Neural Progenitors Persists Long Term with Partial Cytopathic and Limited Immunogenic Effects. Cell Rep. 2016; 15(11): 2315–2322. PubMed Abstract | Publisher Full Text\n\nCugola FR, Fernandes IR, Russo FB, et al.: The Brazilian Zika virus strain causes birth defects in experimental models. Nature. 2016; 534(7606): 267–271. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBell TM, Field EJ, Narang HK: Zika virus infection of the central nervous system of mice. Arch Gesamte Virusforsch. 1971; 35(2): 183–193. PubMed Abstract | Publisher Full Text\n\nFagbami AH: Zika virus infections in Nigeria: virological and seroepidemiological investigations in Oyo State. J Hyg (Lond). 1979; 83(2): 213–219. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFaria NR, Azevedo Rdo S, Kraemer MU, et al.: Zika virus in the Americas: Early epidemiological and genetic findings. Science. 2016; 352(6283): 345–349. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKramer LD, Styer LM, Ebel GD: A global perspective on the epidemiology of West Nile virus. Annu Rev Entomol. 2008; 53: 61–81. PubMed Abstract | Publisher Full Text\n\nLeparc-Goffart I, Nougairede A, Cassadou S, et al.: Chikungunya in the Americas. Lancet. 2014; 383(9916): 514. PubMed Abstract | Publisher Full Text\n\nPAHO: Chikungunya: Statistic Data [Internet]. [cited 25 Jul 2016]. Reference Source\n\nNational Oceanic and Atmospheric Administration: Global analysis—annual 2015 [Internet]. [cited 25 Jul 2016]. Reference Source\n\nKilpatrick MA, Meola MA, Moudy RM, et al.: Temperature, viral genetics, and the transmission of West Nile virus by Culex pipiens mosquitoes. PLoS Pathog. Public Library of Science; 2008; 4(6): e1000092. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRueda LM, Patel KJ, Axtell RC, et al.: Temperature-dependent development and survival rates of Culex quinquefasciatus and Aedes aegypti (Diptera: Culicidae). J Med Entomol. 1990; 27(5): 892–898. PubMed Abstract | Publisher Full Text\n\nDelatte H, Gimonneau G, Triboire A, et al.: Influence of temperature on immature development, survival, longevity, fecundity, and gonotrophic cycles of Aedes albopictus, vector of chikungunya and dengue in the Indian Ocean. J Med Entomol. 2009; 46(1): 33–41. PubMed Abstract | Publisher Full Text\n\nHartley DM, Barker CM, Le Menach A, et al.: Effects of temperature on emergence and seasonality of West Nile virus in California. Am J Trop Med Hyg. 2012; 86(5): 884–894. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaz S, Semenza JC: El Niño and climate change--contributing factors in the dispersal of Zika virus in the Americas? Lancet. 2016; 387(10020): 745. PubMed Abstract | Publisher Full Text\n\nMlakar J, Korva M, Tul N, et al.: Zika Virus Associated with Microcephaly. N Engl J Med. 2016; 374(10): 951–958. PubMed Abstract | Publisher Full Text\n\nReefhuis J, Gilboa SM, Johansson MA, et al.: Projecting Month of Birth for At-Risk Infants after Zika Virus Disease Outbreaks. Emerg Infect Dis. 2016; 22(5): 828–832. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCao-Lormeau VM, Roche C, Teissier A, et al.: Zika virus, French polynesia, South pacific, 2013. Emerg Infect Dis. 2014; 20(6): 1085–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCauchemez S, Besnard M, Bompard P, et al.: Association between Zika virus and microcephaly in French Polynesia, 2013–15: a retrospective study. Lancet. Elsevier; 2016; 387(10033): 2125–2132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSBOC: Federated States of Micronesia [Internet]. [cited 25 Jul 2016]. Reference Source\n\nTsai TF, Popovici F, Cernescu C, et al.: West Nile encephalitis epidemic in southeastern Romania. Lancet. 1998; 352(9130): 767–771. PubMed Abstract | Publisher Full Text\n\nZeller HG, Schuffenecker I: West Nile virus: an overview of its spread in Europe and the Mediterranean basin in contrast to its spread in the Americas. Eur J Clin Microbiol Infect Dis. 2004; 23(3): 147–156. PubMed Abstract | Publisher Full Text\n\nWeinberger M, Pitlik SD, Gandacu D, et al.: West Nile fever outbreak, Israel, 2000: epidemiologic aspects. Emerg Infect Dis. 2001; 7(4): 686. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLanciotti RS, Roehrig JT, Deubel V, et al.: Origin of the West Nile virus responsible for an outbreak of encephalitis in the northeastern United States. Science. 1999; 286(5448): 2333–2337. PubMed Abstract | Publisher Full Text\n\nBrault AC, Huang CY, Langevin SA, et al.: A single positively selected West Nile viral mutation confers increased virogenesis in American crows. Nat Genet. 2007; 39(9): 1162–1166. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhatt S, Gething PW, Brady OJ, et al.: The global distribution and burden of dengue. Nature. 2013; 496(7446): 504–507. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStanaway JD, Shepard DS, Undurraga EA, et al.: The global burden of dengue: an analysis from the Global Burden of Disease Study 2013. Lancet Infect Dis. 2016; 16(6): 712–723. PubMed Abstract | Publisher Full Text\n\nDejnirattisai W, Supasa P, Wongwiwat W, et al.: Dengue virus sero-cross-reactivity drives antibody-dependent enhancement of infection with zika virus. Nat Immunol. 2016. PubMed Abstract | Publisher Full Text\n\nPriyamvada L, Quicke KM, Hudson WH, et al.: Human antibody responses after dengue virus infection are highly cross-reactive to Zika virus. Proc Natl Acad Sci U S A. 2016; 113(28): 7852–7857. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaul LM, Carlin ER, Jenkins MM, et al.: Dengue Virus Antibodies Enhance Zika Virus Infection [Internet]. bioRxiv. 2016; 050112. Publisher Full Text\n\nBarba-Spaeth G, Dejnirattisai W, Rouvinski A, et al.: Structural basis of potent Zika-dengue virus antibody cross-neutralization. Nature. 2016. PubMed Abstract | Publisher Full Text\n\nHalstead SB: Neutralization and antibody-dependent enhancement of dengue viruses. Adv Virus Res. Academic Press, 2003; 60: 421–467. PubMed Abstract | Publisher Full Text\n\nCastro MC: Zika virus and the 2016 Olympic Games - Evidence-based projections derived from dengue do not support cancellation. Travel Med Infect Dis. 2016; pii: S1477-8939(16)30073-4. PubMed Abstract | Publisher Full Text\n\nMassad E, Coutinho FA, Wilder-Smith A: Is Zika a substantial risk for visitors to the Rio de Janeiro Olympic Games? Lancet. 2016; 388(10039): 25. PubMed Abstract | Publisher Full Text\n\nMassad E, Wilder-Smith A, Ximenes R, et al.: Risk of symptomatic dengue for foreign visitors to the 2014 FIFA World Cup in Brazil. Mem Inst Oswaldo Cruz. 2014; 109(3): 394–397. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAguiar M, Rocha F, Pessanha JE, et al.: Carnival or football, is there a real risk for acquiring dengue fever in Brazil during holidays seasons? Sci Rep. 2015; 5: 8462. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLewnard JA, Gonsalves G, Ko AI: Low Risk for International Zika Virus Spread due to the 2016 Olympics in Brazil. Ann Intern Med. 2016. PubMed Abstract | Publisher Full Text\n\nHonório NA, Codeço CT, Alves FC, et al.: Temporal distribution of Aedes aegypti in different districts of Rio de Janeiro, Brazil, measured by two types of traps. J Med Entomol. 2009; 46(5): 1001–1014. PubMed Abstract | Publisher Full Text\n\nMicieli MV, Campos RE: Oviposition activity and seasonal pattern of a population of Aedes (Stegomyia) aegypti (L.) (Diptera: Culicidae) in subtropical Argentina. Mem Inst Oswaldo Cruz. 2003; 98(5): 659–663. PubMed Abstract | Publisher Full Text\n\nGuerbois M, Fernandez-Salas I, Azar SR, et al.: Outbreak of Zika virus infection, Chiapas State, Mexico, 2015, and first confirmed transmission by Aedes aegypti mosquitoes in the Americas. J Infect Dis. 2016; pii: jiw302. PubMed Abstract | Publisher Full Text\n\nAyres CF: Identification of Zika virus vectors and implications for control. Lancet Infect Dis. 2016; 16(3): 278–279. PubMed Abstract | Publisher Full Text\n\nGarcia-Rejon J, Loroño-Pino MA, Farfan-Ale JA, et al.: Dengue virus-infected Aedes aegypti in the home environment. Am J Trop Med Hyg. 2008; 79(6): 940–950. PubMed Abstract\n\nDzul-Manzanilla F, Martínez NE, Cruz-Nolasco M, et al.: Evidence of vertical transmission and co-circulation of chikungunya and dengue viruses in field populations of Aedes aegypti (L.) from Guerrero, Mexico. Trans R Soc Trop Med Hyg. 2016; 110(2): 141–144. PubMed Abstract | Publisher Full Text\n\nMassad E, Tan SH, Khan K, et al.: Estimated Zika virus importations to Europe by travellers from Brazil. Glob Health Action. 2016; 9: 31669. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMessina JP, Kraemer MU, Brady OJ, et al.: Mapping global environmental suitability for Zika virus. eLife. 2016; 5: pii: e15272. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrills A, Morrison S, Nelson B, et al.: Projected Zika Virus Importation and Subsequent Ongoing Transmission after Travel to the 2016 Olympic and Paralympic Games - Country-Specific Assessment, July 2016. MMWR Morb Mortal Wkly Rep. 2016; 65(28): 711–715. PubMed Abstract | Publisher Full Text\n\nD’Ortenzio E, Matheron S, Yazdanpanah Y, et al.: Evidence of Sexual Transmission of Zika Virus. N Engl J Med. 2016; 374(22): 2195–2198. PubMed Abstract | Publisher Full Text\n\nCDC Press Releases [Internet]. 1 Jan 2016 [cited 25 Jul 2016]. Reference Source\n\nDavidson A, Slavinski S, Komoto K, et al.: Suspected Female-to-Male Sexual Transmission of Zika Virus - New York City, 2016. MMWR Morb Mortal Wkly Rep. 2016; 65(28): 716–717. PubMed Abstract | Publisher Full Text\n\nFoy BD, Kobylinski KC, Chilson Foy JL, et al.: Probable non-vector-borne transmission of Zika virus, Colorado, USA. Emerg Infect Dis. 2011; 17(5): 880–882. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHills SL, Russell K, Hennessey M, et al.: Transmission of Zika Virus Through Sexual Contact with Travelers to Areas of Ongoing Transmission - Continental United States, 2016. MMWR Morb Mortal Wkly Rep. 2016; 65(8): 215–216. PubMed Abstract | Publisher Full Text\n\nMcCarthy M: Zika virus was transmitted by sexual contact in Texas, health officials report. BMJ. 2016; 352: i720. PubMed Abstract | Publisher Full Text\n\nTowers S, Brauer F, Castillo-Chavez C, et al.: Estimation of the reproduction number of the 2015 Zika virus outbreak in Barranquilla, Colombia, and a first estimate of the relative role of sexual transmission [Internet]. arXiv. 2016. Reference Source\n\nLazear HM, Diamond MS: Zika Virus: New Clinical Syndromes and Its Emergence in the Western Hemisphere. J Virol. 2016; 90(10): 4864–4875. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCDC: Guidelines for Travelers Visiting Friends and Family in Areas with Chikungunya, Dengue, or Zika [Internet]. [cited 25 Jul 2016]. Reference Source\n\nThe Lancet Infectious Diseases: Zika virus at the games: is it safe? Lancet Infect Dis. 2016; 16(6): 619. PubMed Abstract | Publisher Full Text\n\nMcConnell J, de Ambrogi M, Cleghorn S, et al.: Zika virus and the 2016 Olympic Games - Editors’ reply. Lancet Infect Dis. Elsevier, 2016; pii: S1473-3099(16)30266-3. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "16240",
"date": "12 Sep 2016",
"name": "Eduardo Massad",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn an attempt to justify the disastrous Iraq invasion in 2002, the then 13th Secretary of Defense of the USA, Donald Rumsfeld, stunned reporters when he uttered his (in)famous reflection on \"known knowns\", \"known unknowns\" and \"unknown unknowns\" (Rumsfeld, 2011).\nIn their interesting review on Zika epidemics, Grubauch and Andersen address the panic ensued by the overwhelmed number of cases in Latin America and elsewhere. They call the attention to several \"known knowns\" and \"known unknowns\" of the current Zika epidemics, although they do not shy away of pointing to possible \"unknown unknowns\".\nGrubauch and Andersen find no conclusive evidence in the literature for significant biological differences between the Latin American Zika virus strain and those circulating elsewhere. They argue that the epidemic scale in the Americas has been influenced by climate, humans and mosquitoes population densities and local prevalence of other viruses, in particular flaviviruses, like dengue and yellow fever. The severe neurological abnormalities associated with Zika virus would not be a new trait of the Latin America strain, but rather may have been overlooked due to previously small outbreaks.\nEvery time the world is hit by an emergent or re-emergent pathogen with pandemic potential, panic ensues (Amaku et al., 2016). This is understandable due to previous history of recurrent pandemics, like the medieval Black Death or the 1918 Spanish Flu, which killed millions of people in Europe and around the world (Massad et al., 2007; Burattini, Coutinho & Massad, 2009). After the Pasteur-Koch germ theory was proposed, the previously \"unknown unknowns\" could explain the panic. The recent panic observed when SARS, H5N1 and more recently MERS-CoV emergence, or the current Zika, should not be justified due to the prompt discovery of the ultimate cause and the possible control mechanisms. Unlike Londoners of the XVII century (Bell, 2001), who burnt witches and culled thousands of cats (by the way, both excellent rat killers), we have the scientific tools and mechanisms to face the threat with the necessary rationality. In the case of Zika virus, however, many \"known unknowns\" still remain, like the true phenotypic repercussion of genotypic differences between the strains circulating in different parts of the world, differences in mosquito competences, the actual number of people that has already been infected, the competition by the vector of different viruses transmitted by the same aedes mosquitoes, vertical transmission of the virus in hosts and vectors, neuropathogenic potential of different strains, cross-immunity, antibody-enhancement by previous infection, just to mention a few. In addition, it is not well known the actual role of sexual transmission of Zika in triggering or maintaining an outbreak. As for the \"unknown unknowns\", just time will disclose.\nOne important unknown currently being discussed in the literature is whether Zika will disappear from the affected regions and whether and when it will recur. Mathematical models, very useful in the understanding and control of previous epidemics, have been widely applied in an attempt to describe and make prediction about the current Zika outbreak (Massad et al., 2016)). We now know that the Basic Reproduction Number, R0, of Zika is around 3, slightly higher than that of Dengue (Coelho et al., 2008) and it is even possible to predict the likelihood of Zika being exported to unaffected regions of the world, either causing a single and self limited outbreak or establishing itself, depending on whether localis lesser or greater than 1.\nFor the sake of completeness, two additional speculative unknowns are worth mentioning. The circulation of Zika virus and its potential interaction with dengue raises new concerns regarding vaccination strategies against the latter. The subtle trade-off justifying the recommendation of the vaccine might no longer hold. On the other hand, the patchy distribution of serious outcomes due to Zika has not been satisfactorily explained. Our navigation map cannot overlook the apparent clustering of cases of microcephaly reported in northeastern Brazilian states so far.\nWe think that the many \"known unknowns\" related to Zika epidemic explored in this paper are worthwhile being published and we are sure that Grubauch and Andersen reflections could help bring science and rationality to the fore, soothing the perhaps unjustifiable panic. Althoug Grubauch and Anderson do not cross all the t's nor dot all the i's, they at least show us all the uncrossed t's and undotted i's of the current Zika \"unknowns\" epidemic. This definitively qualify their paper for publication.\nFinally, for the record: the World Health Organization officially declared - zero cases of Zika during the Olympic Games!",
"responses": [
{
"c_id": "2182",
"date": "12 Sep 2016",
"name": "Nathan Grubaugh",
"role": "Author Response",
"response": "Fantastic review! Thank you for taking the time and pointing out some of the issues that we did not get into, mainly because we were already >4000 words! There are so many unknowns of many varieties, we just ask for over-speculation to be tempered until the science can catch up. No new Zika virus infections during the Olympics is amazing news!"
}
]
},
{
"id": "16130",
"date": "23 Sep 2016",
"name": "Gonzalo Moratorio",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this interesting and well-executed review, Grubauch and Andersen 'navigate' the literature to provide a clean, well-driven and totally enjoyable text about the current situation generated for the Zika outbreak in the Americas. Based on the bibliography, they conclude there is no scientific evidence that the Brazilian Zika strain presents a higher pathogenicity compared to others circulating elsewhere. In addition, as was discussed in previous reports, they agreed with the role played by urban and weather factors to particularly enhance this outbreak. The authors also believe that is totally necessary to confirm the connection between Dengue and Zika as a risk factor in disease severity. Finally, given the sample size of the recent outbreak, we could have under estimated rare diseases associated to this virus in the past.\n\nThe article itself is very nicely written and provides a very balanced viewpoint of Zika virus, which is something that has been lacking from several media sources. The authors have put significant thought into the attributes that affect the severity of the outbreak, but perhaps the most important statements concerning the evolution of the virus and the neurological phenotypes recently observed. With their calculations for the number of microcephaly cases that could have been detected on Yap Island (quite possibly none due to the size of the outbreak), they suggest that statistics and surveillance, rather than genetic differences, affect disease severity. Of course, additional research is necessary into understanding the severity of the Zika virus outbreak and its connections to microcephaly and Guillain-Barre syndrome and whether the genetic differences between the strains of Zika virus are or are not responsible for these newly observed phenotypes.\n\nThe authors also expand on the details on how Zika virus could infect a traveler in Brazil and then induce an outbreak in their home country, which they estimate to have a much lower probability than popular media sources might suggest. Though the commentary surrounding the possibility of the Olympics enhancing Zika virus comes after the games, the thoughtful consideration of the risks supplies rational thinking that has been lacking, especially when juxtaposed with the hysteria prior to the games. The review superbly quells the hyperbole that has surrounded the Zika outbreak in the Americas. Regardless of this low possibility of traveler-associated transmission of Zika virus in new locales, regions with the proper conditions for Zika virus transmission should remain vigilant, and continued campaigns of mosquito control, especially given the breadth of viruses spread by mosquitoes (including dengue, chikungunya and West Nile viruses).\n\nWe would like to highlight that the authors didn’t cite any of their previous works, which surprised us in a positive way, showing the author’s will to deliver a clean and not biased opinion about the topic.\n\nMinor comments:\nThe article is written for any kind of reader, which we found excellent, but in that case would be nice to clarify what is El Niño. Even if wouldn’t be necessary for most of the readers.\n\nThe paragraph that starts with “One main reason behind …August is winter in Brazil…”\nBrazil is huge, is the he largest country in South America and in the Southern Hemisphere. Is placed 5th in the list sovereign states and dependencies by area in the world. Particularly, the weather in Rio, still in august, can be warm enough for mosquitos. Last August the lower temperature registered was 71 F and the higher 78 F.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1914
|
https://f1000research.com/articles/5-1910/v1
|
03 Aug 16
|
{
"type": "Research Article",
"title": "Prevalence of hepatitis B and C infections in hemodialysis patients",
"authors": [
"Karina Salvatierra",
"Hector Florez",
"Hector Florez"
],
"abstract": "Introduction: Infections with hepatitis B and C viruses (HBV and HCV) are a major global health problem. Patients with chronic renal failure (CRF) on hemodialysis constitute a population at risk of HBV and HCV infections. Objective: Determining prevalence of the surface antigen of hepatitis B virus (HBsAg) and antibodies to hepatitis C virus (anti-HCV) in patients who attended dialysis units in the city of Posadas (Argentina). Materials and methods: The studied population comprised 172 patients with CRF in hemodialysis. HBsAg and anti-HCV antibodies were evaluated by enzyme-linked immunosorbent assay (ELISA). Results: On a total of 172 hemodialysis patients included in the study, 98 were males (57%) and 74 were females (43%), aged between 12 and 85 years (mean 53.4). 8.7% (15/172) of the patients were positive for HBsAg and 9.9% (17/172) were positive for anti-HCV reagents. 72.1% of patients had a hemodialysis treatment time of less than 5 years. A history of having received previous transfusions was observed in both HBsAg positive cases (7/15) and the anti-HCV positive cases (5/17). Elevated transaminase levels were observed in patients with positive and negative serology. Conclusion: The results of this study demonstrate a high prevalence of serological markers for HBV and HCV in patients with CRF on hemodialysis in city of Posadas (Argentina), as compared to cities in developed countries.",
"keywords": [
"hepatitis B",
"hepatitis C",
"chronic kidney failure",
"hemodialysis",
"serological markers"
],
"content": "Introduction\n\nInfections with hepatitis B and C viruses (HBV and HCV, respectively) are a major global health problem affecting 240 million people who suffer from chronic HBV infection and about 150 million who suffer from HCV infection1,2. In most cases, these viruses cause chronic infection whose natural course leads to liver cirrhosis, liver failure and/or hepatocarcinoma in affected patients3.\n\nPatients with chronic renal failure (CRF) on hemodialysis are at high risk of contracting viral infections with HBV and HCV, the most common cause of liver disease in these patients4,5. Therefore, strict procedures for the control of hepatitis must be introduced in all dialysis units6.\n\nThe geographical distribution of HBV infection is not uniform throughout the world. Depending on the prevalence, different areas are classified as high, intermediate or low endemicity. HBV infection is highly prevalent (8–15%) in Southeast Asia, China, the Philippines, Africa, the Amazon basin and the Middle East. In Eastern Europe, Central Asia, Japan, Israel and Russia the prevalence is intermediate (2.7%), while in North America, Western Europe, Australia and South America the prevalence is low (<2%)7. In Latin America it ranges from 2 to 7%8. In developed countries the prevalence of HBV in patients treated with hemodialysis is 1%9, while in developing countries the prevalence ranges from 2% to 20%4,10,11.\n\nThe prevalence of HCV in hemodialysis patients ranges from 2.6 to 22.9% (mean 13.5%) in developed countries, but can reach up to 70% in developing countries12–15. In Latin America, the prevalence of HCV is also highly variable in hemodialysis patients, even within the same country. In Mexico, the prevalence is 6.7%16, in Colombia it ranges from 2.9 to 42.2%17,18, while in Brazil ranges from 6–72% (mean 52%)19.\n\nIn recent years there was a significant decrease in the prevalence of both HBV and HCV infections in industrialized countries20. This decline is attributed, among other factors, to the decrease in transfusions, vaccination against HBV and introduction of general biosecurity measures to prevent transmission of infection in hemodialysis units.\n\nThe aim of this study was to evaluate the prevalence of HBV and HCV in hemodialysis population dialysis units of four hemodialysis centers in the city of Posadas (Argentina).\n\n\nMethods\n\nIn this study, a total of 172 patients diagnosed with CRF under hemodialysis attending four hemodialysis centers of the city of Posadas (Argentina) were included. Ethical approval to conduct the study was obtained by the hemodialysis centres involved in Posada, Argentina: Instituto Privado de Nefrologia srl Roque Saenz Peña I; Instituto Privado de Nefrologia srl Roque Saenz Peña II, Instituto de Nefrologia IOT (Instituto de Ortopedia, Traumatología y Medicina Laboral de alta complejidad); Instituto de Nefrologia Boratti. Institutional Review Board approval from the University of Misiones was not required as the private centers involved are not affiliated with the University and the study was considered a human health research without risks. The protection and control mechanisms for this type of research in Argentina includes inclusion of a written informed consent obtained from each participant\".\n\n: Patients who participated in the study went daily to a hemodialysis center and participation was voluntary. Inclusion criteria encompassed individuals with CRF who were just starting hemodialysis treatment 30 days before obtaining blood samples.\n\nAll patients signed an informed consent statement for joining the study after explaining the scope of the study. A unique and anonymous code was assigned to each patient and patient confidentiality was ensured.\n\nBlood samples were collected through venipuncture or by finger/heel stick in dry tubes (9.5 ml) with vacuum. After clot retraction, samples were centrifuged at 1,500 rpm for 5 minutes and stored in 5mL aliquots at -20°C (http://www.cdc.gov/measles/lab-tools/serology.html).\n\nSerological markers: surface antigen of hepatitis B virus (HBsAg) and antibodies to hepatitis C virus (antiHCV) were detected by enzyme-linked immunosorbent assay (ELISA) using commercially available kits (Wiener, Lab. S.A.I.C.).\n\nThe following information from medical records were obtained: age, gender, time on hemodialysis, alanine aminotransferase (ALT) index, history of transfusions and history of drug abuse. An online information system was developed to analyze the information (available at http://bioinf.itiud.org/hemodialysis.php). The system allows computational analysis and deployment of information through generation of reports and statistics charts, designed to provide a simple reading of the results of the study. A descriptive analysis was performed by calculating means, medians and frequencies. Chi-square test was used to analyze the significant relationships between categorical data and Mann–Whitney U test was used for the comparison of continuous data (p<0.005). Data were managed and analyzed using InfoStat software.\n\n\nResults\n\nOur final sample consisted of 172 patients: 98 males (57%) and 74 females (43%). Their ages ranged from 12 to 85 years with a mean age of 53.4 years (Dataset 1).\n\nAll patients had not history of tattooing, piercing, or use illegal drugs, and no human immunodeficiency virus (HIV) coinfection was found in any of the cases.\n\nThe cause of CRF in patients on hemodialysis was unknown in 31.4% of the cases, followed by diabetic nephropathy (22.1%). The etiology of CRF for these patients is detailed in Figure 1.\n\nSerum samples from hemodialysis patients were tested for presence of HBsAg and antiHCV by ELISA. Fifteen (15/172) cases were positives for HBsAg, whereas 17 (17/172) cases were positives for antiHCV. Figure 2 shows the distribution of HBsAg and antiHCV cases by age group.\n\nThe majority of HBsAg positive cases were females (9/15), but this was not statistically significant (p = 0.16), while of the majority of cases positive for HCV markers were males (14/17); this result was statistically significant (p = 0.02). Figure 3 presents the distribution of HBsAg and antiHCV markers of by gender.\n\nChi-square Pearson test. HBsAg: p = 0.16; antiHCV: p = 0.02.\n\nAll patients participating in this study had 4-hour hemodialysis sessions three times a week and 72.1% of patients showed a mean time of hemodialysis less than 5 years (Figure 4). The duration of hemodialysis was not a significant risk factor for HBsAg (p = 0.9) and antiHCV (p = 0.2) presence.\n\nChi-square Pearson test. HBsAg: p = 0.9; antiHCV: p = 0.2.\n\nTransfusion history was observed in 47% (7/15) of cases positives for HBsAg. Positivity for HBsAg significantly correlated with transfusion (p = 0.003), while 29.4% (5/17) of cases positives for antiHCV did not show correlation with transfusion (p = 0.23) (Figure 5).\n\nChi-square Pearson test. HBsAg: p = 0.003; antiHCV: p = 0.23.\n\nIn our patient population, 154 (89,5%) cases showed normal concentration of serum ALT level (< 40 U/L), and 18 (10,5%) showed elevation in serum ALT level (>40 IU/L) (Figure 6).\n\nMann-Whitney U test. HBsAg: p = 0.001; antiHCV: p = 0.002.\n\nA total of six HBsAg positive cases out of 15 (33.3%) showed elevation in serum ALT level, whereas one (5.5%) of the cases positive for HCV out of 17 showed increased serum ALT level. This correlation was statistically significant (p = 0.001 and p = 0.002 respectively).\n\n\nDiscussion\n\nPatients with CRF on hemodialysis are a group at high risk for HBV and HCV infections.\n\nIn Argentina the prevalence of HBsAg varies between the provinces, from 0.17% to 1.79%, and prevalence of antiHCV varies between 1.70% and 21%21. In our study we found a relatively high prevalence of serological markers for HBV and HCV in hemodialysis patients in Posadas (Argentina). The HBsAg antibody (indicator of active HBV infection) was found in 8.7% of the studied cases and antiHCV in 9.9%.\n\nThese values are comparable to values reported in several studies on hemodialysis patients. In Peru, HBsAg prevalence ranges from 0.2% to 4.8%22, in Brazil varies between 4% and 28%23,24, in Cuba is 4%25, and in Colombia HBsAg is 22%26.\n\nVarious studies in hemodialysis units show different prevalence rates for HCV: in Peru between 90% and 4.65%22, in Uruguay from 16% to 3%27, in Brazil 33.4% to 16%23,28, in Mexico 6.7%16, and in Cuba 90%25. The antiHCV prevalence in dialysis units in Cali (2.9%) and Bogotá (2.7%) is very low17,29, while in Medellin is high (42,2%)18.\n\nBy exploring the correlation between risk factors and seropositivity for HBV and HCV in hemodialysis patients, it seemed that timing of hemodialysis was not a risk factor. History of transfusion correlated with the risk of HBV infection, but not with risk of HCV infection. The elevation of ALT enzyme activity was not a good index for HBV and HCV infection in these patients, since normal values were observed in a high percentage of patients positive for HBsAg and antiHCV.\n\n\nConclusion\n\nThe results of this study demonstrate a high prevalence of serological markers for HBV and HCV infections in CRF patients on hemodialysis in Argentina, compared with results obtained from patients in developed countries.\n\nTherefore, an effective strategy to prevent nosocomial transmission of HBV and HCV in hemodialysis units and reduce the prevalence of infection should be implemented in strict compliance with biosafety standards, measures of education, hygiene and HBV vaccination plans to prevent the infection.\n\n\nConsent\n\nWritten informed consent to participate in the study and publish clinical data was obtained by the patients.\n\n\nData availability\n\nF1000Research: Dataset 1. Patient characteristics, 10.5256/f1000research.9068.d13138530",
"appendix": "Author contributions\n\n\n\nKarina Salvatierra: study design, sample processing, data analysis, discussion of results, and writing the manuscript. Hector Florez: design and analysis of data, drafting the manuscript. Both authors agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work has been supported by the Information Technologies Innovation (ITI) Research Group.\n\n\nAcknowledgments\n\nThe authors would like to thank the hemodialysis patients, authorities, and the staff of the four dialysis centers that collaborated in the study, making this research possible. Both authors agreed to the final content of the article.\n\n\nReferences\n\nMcMahon BJ: Epidemiology and natural history of hepatitis B. Semin Liver Dis. 2005; 25(Suppl 1): 3–8. PubMed Abstract | Publisher Full Text\n\nShepard CW, Finelli L, Alter MJ: Global epidemiology of hepatitis C virus infection. Lancet Infect Dis. 2005; 5(9): 558–67. PubMed Abstract | Publisher Full Text\n\nPerz JF, Armstrong GL, Farrington LA, et al.: The contributions of hepatitis B virus and hepatitis C virus infections to cirrhosis and primary liver cancer worldwide. J Hepatol. 2006; 45(4): 529–38. PubMed Abstract | Publisher Full Text\n\nVladutiu DS, Cosa A, Neamtu A, et al.: Infections with hepatitis B and C viruses in patients on maintenance dialysis in Romania and in former communist countries: yellow spots on a blank map? J Viral Hepat. 2000; 7(4): 313–19. PubMed Abstract | Publisher Full Text\n\nOtedo AE, Mc’Ligeyo SO, Okoth FA, et al.: Seroprevalence of hepatitis B and C in maintenance dialysis in a public hospital in a developing country. S Afr Med J. 2003; 93(5): 380–84. PubMed Abstract\n\nMallick NP, Gokal R: Haemodialysis. Lancet. 1999; 353(9154): 737–42. PubMed Abstract | Publisher Full Text\n\nCDC: Hepatitis B FAQs for Health Professionals. 2011. Reference Source\n\nTanaka J: Hepatitis B epidemiology in Latin America. Vaccine. 2000; 18(Suppl 1): S17–S19. PubMed Abstract | Publisher Full Text\n\nTokars JI, Frank M, Alter MJ, et al.: National surveillance of dialysis-associated diseases in the United States, 2000. Semin Dial. 2002; 15(3): 162–71. PubMed Abstract | Publisher Full Text\n\nBoulaajaj K, Elomari Y, Elmaliki B, et al.: [Prevalence of hepatitis C, hepatitis B and HIV infection among haemodialysis patients in Ibn-Rochd university hospital, Casablanca]. Nephrol Ther. 2005; 1(5): 274–84. PubMed Abstract | Publisher Full Text\n\nTeles SA, Martins RM, Gomes SA, et al.: Hepatitis B virus transmission in Brazilian hemodialysis units: serological and molecular follow-up. J Med Virol. 2002; 68(1): 41–49. PubMed Abstract | Publisher Full Text\n\nMeyers CM, Seeff LB, Stehman-Breen CO, et al.: Hepatitis C and renal disease: an update. Am J Kidney Dis. 2003; 42(4): 631–657. PubMed Abstract | Publisher Full Text\n\nBergman S, Accortt N, Turner A, et al.: Hepatitis C infection is acquired pre-ESRD. Am J Kidney Dis. 2005; 45(4): 684–9. PubMed Abstract | Publisher Full Text\n\nBiamino E, Caligaris F, Ferrero S, et al.: [Prevalence of anti-HCV antibody positivity and seroconversion incidence in hemodialysis patients]. Minerva Urol Nefrol. 1999; 51(2): 53–55. PubMed Abstract\n\nJadoul M, Poignet JL, Geddes C, et al.: The changing epidemiology of hepatitis C virus (HCV) infection in haemodialysis: European multicentre study. Nephrol Dial Transplant. 2004; 19(4): 904–9. PubMed Abstract | Publisher Full Text\n\nMéndez-Sánchez N, Motola-Kuba D, Chavez-Tapia NC, et al.: Prevalence of hepatitis C virus infection among hemodialysis patients at a tertiary-care hospital in Mexico City, Mexico. J Clin Microbiol. 2004; 42(9): 4321–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamírez R, Fernández J, Guevara JG, et al.: Prevalencia de anticuerpos contra el virus de hepatitis C en unidades de diálisis de Cali-Colombia. Rev Col Gastroenterol. 2010; 25(1): 14–8. Reference Source\n\nEchavarría E: Estudio de anticuerpos contra el virus de la hepatitis C en donantes de sangre y grupos de alto riesgo. Acta Médica Colombiana. 1992; 17: 11–15. Reference Source\n\nMedeiros MT, Lima JM, Lima JW, et al.: Prevalence and associated factors to hepatitis C in hemodialysis patients in Brazil. Rev Saude Publica. 2004; 38(2): 187–193. PubMed Abstract | Publisher Full Text\n\nAlashek WA, McIntyre CW, Taal MW: Hepatitis B and C infection in haemodialysis patients in Libya: prevalence, incidence and risk factors. BMC Infect Dis. 2012; 12: 265. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarinovich S, Lavorato C, Bisigniano L, et al.: Registro Argentino de Diálisis Crónica SAN-INCUCAI 2011. Informe 2012. Revista Nefrología Argentina. 2013; 11(2). Reference Source\n\nLoza Munárriz C, Depaz Dolores MY, Suarez Jara M, et al.: Frecuencia de marcadores serológicos de hepatitis viral B y C en pacientes que ingresan por primera vez al programa de hemodiálisis en el Hospital Nacional Cayetano Heredia. Rev Gastroenterol Perú. 2005; 25(4): 320–327. Reference Source\n\nSouza KP, Luz JA, Teles SA, et al.: Hepatitis B and C in the hemodialysis unit of tocantins, brazil: serological and molecular profiles. Mem Inst Oswaldo Cruz. 2003; 98(5): 599–603. PubMed Abstract | Publisher Full Text\n\nCarrilho FJ, Moraes CR, Pinho JR, et al.: Hepatitis B virus infection in Haemodialysis Centres from Santa Catarina State, Southern Brazil. Predictive risk factors for infection and molecular epidemiology. BMC Public Health. 2004; 4: 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartínez Córdova Z, Peña Fresneda N: Prevalencia de anticuerpos anti-VHC y del antígeno de superficie de la hepatitis B en pacientes tratados con hemodiálisis. Rev Cubana Med. 2008; 47: 1. Reference Source\n\nBeltrán M, Navas MC, Arbeláez MP, et al.: Seroprevalencia de infección por virus de la hepatitis B y por virus de la inmunodeficiencia humana en una población de pacientes con múltiples transfusiones en cuatro hospitales, Colombia, Sur América. Biomédica. 2009; 29: 232–43. Publisher Full Text\n\nDe la Hoz F: Epidemiología de la hepatitis C en Latinoamérica y Colombia. Revista Repertorio de Medicina y Cirugía. 2002; 11(1).\n\nMoraes CR, Carrilho FJ, Bassit LC, et al.: Hepatitis C virus infection in hemodialysis patients from southern Brazil. Epidemiological data and genotypes. Hepatology. 2000; 32(4 pt 2): 546A.\n\nManascero-Gómez AR, Gutiérrez MF, Jaramillo-Tobón C, et al.: Evaluación de algunos factores de riesgo de contraer hepatitis C en pacientes hemodializados. Universitas Scientiarum. 2007; 12(III): 47–56. Reference Source\n\nSalvatierra K, Florez H: Dataset 1 in: Prevalence Analysis of hepatitis B and C in hemodialysis patients. F1000Research. 2016. Data Source"
}
|
[
{
"id": "15420",
"date": "22 Aug 2016",
"name": "Marina C Berenguer",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe aim of this study was to evaluate the prevalence of HBV and HCV in hemodialysis population dialysis units of four hemodialysis centers in the city of Posadas (Argentina).\n\nSerum samples from hemodialysis patients were tested for presence of HBsAg and antiHCV by ELISA. Fifteen (15/172) cases were positives for HBsAg, whereas 17 (17/172) cases were positives for antiHCV.\n\nThis is a descriptive study on local seroprevalence of HBV and HCV infection in a region of Argentina. The title is appropriate. The abstract provides adequate summary of the content. The methods are adequately described. The conclusions are adequate.\nThere is though missing information which is relevant: no information on viremia, underlying disease severity, sequential data overtime.\nThe following questions should be asked to the authors:\nHow many of the patients were viremic? Was there any information on underlying disease severity How many were HBV and HCV coinfected? Do the authors have sequential data to try to clarify whether there is an increase, stability or decrease of positive cases? Were there HBV patients coinfected with HDV? Were there differences in positivity results by center?\nIn essence, this is a simple study describing prevalence of HBV and HCV seromarkers in a particular region of Argentina. There have been many studies published to date on seroprevalence of these viruses around the world.",
"responses": [
{
"c_id": "2166",
"date": "30 Aug 2016",
"name": "karina salvatierra",
"role": "Author Response",
"response": "Dear Reviewer Marina C Berenguer Thank you so much for your feedback. We want to inform you: - Unfortunately, we do not have information regarding viremic, disease severity, and sequential data. - There were not patients coinfected with with HBV and HCV. - This is a simple study describing prevalence of HBV and HCV seromarkers in a particular region of Argentina. So far, there is no studies regarding this topic in this region Given the limitations on what is counted in the region to do research, this is the first step to go ahead and propose further research. Thank you so much Best Regards"
}
]
},
{
"id": "17997",
"date": "05 Dec 2016",
"name": "Isabella Esposito",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPatients with chronic renal failure (CRF) receiving hemodialysis are at higher risk for acquiring Hepatitis B Virus (HBV) and Hepatitis C Virus (HCV) infections than the general population. The susceptibility to acquire viral hepatitis during hemodialysis has several potential underlying reasons related to both the patient and the hemodialysis procedure. Strict infection control measures are essential to prevent nosocomial transmission.\nSo the aim of this article was to investigate the prevalence of HBV and HCV infection in the hemodialysis population of a region of Argentina as well as risk factors for infection.\nSerum testing for the presence of hepatitis B surface antigen (HBsAg) and anti-HCV antibodies (antiHCV) by ELISA and subsequent correlation of sero-positivity to HBV and HCV with timing of hemodialysis, history of transfusion, alanine aminotransferase (ALT) index, allowed the authors the opportunity to describe HBV and HCV prevalence in the present cohort of hemodialysis patients.\nThe study is adequately described; however, it could have been more detailed. Some suggestions are as follow:\nPlease compare the difference in sero-positivity results between the four hemodialysis centers in the city of Posadas, which can eventually reflect difference in the biosecurity procedures of each center.\n\nDescribe whether the patients included in the study were being vaccinated against hepatitis B virus.\n\nIt would be necessary to clarify the HBV and HCV serological status of patients with CRF before starting hemodialysis treatment.\n\nIn the Methods section, please state the period of time of sample collection.\n\nThe authors correlated transfusion history with HBsAg or anti-HCV serological status and reported a significant association with positivity for HBsAg. It would be important to indicate the year of transfusion, in order to rule out (if it was a recent event) or strongly suggest (if it occurred before the 1960s) that transfusion was the source of HBV infection.\n\nIt would have been interesting to determine anti-HBc total antibodies as well.\n\nThere is potential for more in-depth analysis and comparison. The authors should include more discussion and compare with their findings. For example, it is known that Misiones is a high HBV prevalence area in Argentina (please read the papers by Delfino et al., J ClinVirol 2012 and Arch Virol 2014).\nThis is a well written but very simple study describing prevalence of HBV and HCV seromarkers in a particular region of Argentina. It could be updated with more detailed information and we do not consider it to be of an acceptable scientific standard. Therefore, this article is not suitable unless the content is extensively edited.",
"responses": []
},
{
"id": "15421",
"date": "21 Dec 2016",
"name": "Rocio Hassan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a descriptive work that has as primary aim to evaluate the prevalence of HBV and HCV in a hemodialysis population of the city of Posadas (Argentina).\nThe research have been conducted with ethical and methodological correctness and results, although of limited reach, contribute with useful information on HBV and HCV prevalence in an at-risk group (hemodialysis patients) from a geographical region with paucity of such studies. Moreover, sero-prevalence of HBV in Misiones Province appear to be high, which may indicate a distinct epidemiological situation closer to that reported for neighbor Brazil than that described for Central and Southern Argentina.\nTherefore, I consider that the manuscript is suitable for indexing, provided some restructuring is done in the text. Moreover, the article has, at present, 147 views and 50 download. This means it has raised interest in the scientific community and for that, I think it deserves to be corrected to include the suggestions of the reviewers.\nI would like, then, to make some suggestions that I think may help improving the text.\nIn first place, and if it is possible, the present title “Prevalence of hepatitis B and C infections in hemodialysis patients” may be completed to include a characterization of the geographical region from where the results come from. The reader would then have an insight of the specific geographic and epidemiological context from the beginning.\n\nIn the introduction, it will be very important to geographically and epidemiologically situate Misiones Province, and specifically Posadas City, describing available results on HBV/HCV prevalence in this specific population.\n\nMethods: It is desirable to check the English syntax and grammar of the first to third paragraphs. The phrase “The protection and control mechanisms for this type of research in Argentina includes inclusion of a written informed consent obtained from each participant\" is unnecessary, since in the third paragraph it is stated that “All patients signed an informed consent statement”.\n\nPlease, explain better the inclusion criteria. The statement “individuals with CRF who were just starting hemodialysis treatment 30 days before obtaining blood samples” is misleading, since it gives the impression that all individuals started hemodialysis 30 days before obtaining blood samples for the present study.\n\nIn the statistics description, replace “Mann–Whitney U test was used for the comparison of continuous data (p<0.005)” by Mann–Whitney U test was used for the comparison of continuous non-normal data. P-values <0.05 were considered significant.\n\nResults: Please clarify the issue of HIV infection (vs. co-infection). Have all patients been tested for HIV infection, or only the HCV/HBV positive ones? If it is the first case, please correct the statement “human immunodeficiency virus (HIV) coinfection was found in any of the cases.” by “human immunodeficiency virus (HIV) infection was not found in any of the cases.”\n\nFigure 1, please express decimal notation with periods instead of commas.\n\nGender association in HCV+ and HBV+ patients would be better presented as a stacked column graphic with percentage of male or female individuals for each category. Wouldn’t Fisher’s exact test be a better statistical to apply here?. The authors should provide in the Discussion a possible explanation for the gender association found in the study.\n\nWhat was the specific statistical approach used for testing association between time of dialysis and HCV/HBV serostatus? Was it, as described in Fig. 4, a 2x6 Pearson Chsq test or, alternatively time was categorized in >5 vs < 5 years? Risk is not well modelled by Chsq tests, and so, with the present statistical approach it only can be described as association. To model the risk, a logistic regression with HCV/HBV serostatus as dependent variable may be used instead. Again, the bar graphic with number of pts as scale is not very informative.\n\nRespect of data on serum ALT levels, in figure 6, results of a Mann-Whitney test are provided, but the representation of data is as categories (till 40U/L - >40 U/L). The authors should either present categorical test results (i.e. Chisq or Fisher’s) or a Box-and-whisker plot with data around the median value. This last would be a better graphical depiction for the data.\n\nDiscussion needs to be rewritten to put the results in a stronger theoretical context. The authors should focus at least part of the discussion on the comparison of Misiones data with Brazilian available data and data from the rest of Argentina. In fact, most of the significant references from Brazil are listed in the article reference list, but they have not been used to highlight the similarity of Misiones’ data. Important references from Misiones (and the rest of Argentina) are lacking, as it was pointed out in a previous review.\n\nAdditionaly, I would like to ask the authors to include in the revised version of the ms some points addressed by a previous reviewer, which in my opinion are minor revisions.\nAdd information on demographics and the prevalence of HCV and HBV separately for each of the four participating hemodialysis centers in the city of Posadas. The data can be included as a supplementary table, if so allowed by the Journal.\n\nPlease, describe the HBV vaccine history of patients.\n\nIf possible, include the information of HBV and HCV serological status of CRF patients prior to hemodialysis.\n\nIn the Methods section, please state the period of time of sample collection.\n\nIn respect to the association between transfusion history and HBsAg, it would be important to indicate the period of transfusion, in order to rule out (if it was a recent event) or strongly suggest (if it occurred before the 1960s) that transfusion was the source of HBV infection.\n\nThere is potential for more in-depth analysis and comparison. The authors should include more discussion and compare with their findings. For example, it is known that Misiones is a high HBV prevalence area in Argentina (please read the papers by Delfino et al., J ClinVirol 2012 and Arch Virol 2014).",
"responses": []
},
{
"id": "15836",
"date": "21 Mar 2017",
"name": "Rafaela Ferrari",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nEven though it is a descriptive research, I think it's important for local researchers. However, I suggest some changes:\nthe title “Prevalence of hepatitis B and C infections in hemodialysis patients” should be completed to include the geographical region\n\nP-values <0.05 were considered significant, not <0.005, is possible there were a typing error\n\nHow many of the patients presented the symptomatology of hepatitis?\n\nHow many of the patients included in the study were being vaccinated against hepatitis B virus?",
"responses": [
{
"c_id": "3073",
"date": "02 Oct 2017",
"name": "karina salvatierra",
"role": "Author Response",
"response": "Dear Reviewer Rafaela Ferrari. Thank you so much for your feedback. We want to inform you: 1. In the abstract and in the text we describe the geographic location. 2. We have changed the P-values as you suggested. Mann–Whitney U test was used for the comparison of continuous non-normal data. P-values <0.05 were considered significant. 3. Hepatitis C is usually asymptomatic it rarely causes symptoms. Hepatitis C may not produce symptoms. The spectrum of the symptomatology of hepatitis B disease varies from subclinical hepatitis to icteric hepatitis to fulminant, acute, and subacute hepatitis during the acute phase, and from an asymptomatic chronic infection state to chronic hepatitis, cirrhosis, and hepatocellular carcinoma (HCC) during the chronic phase. In this case, patients was not presented symptomatology of hepatitis. 4. In respect of hepatitis B vaccination, 125/172 (73%) of the patients completed the vaccination scheme, while none of the HBsAg (+) cases have had HBV vaccination. Thank you so much Best Regards"
}
]
}
] | 1
|
https://f1000research.com/articles/5-1910
|
https://f1000research.com/articles/5-1680/v1
|
13 Jul 16
|
{
"type": "Case Report",
"title": "Case Report: ALCAPA syndrome: successful repair with an anatomical and physiological alternative surgical technique",
"authors": [
"Luis Gustavo Vilá Mollinedo",
"Andrés Jaime Uribe",
"José Luis Aceves Chimal",
"Roberto Pablo Martínez-Rubio",
"Karen Patricia Hernández-Romero",
"Andrés Jaime Uribe",
"José Luis Aceves Chimal",
"Roberto Pablo Martínez-Rubio",
"Karen Patricia Hernández-Romero"
],
"abstract": "Anomalous left coronary artery from the pulmonary artery, or ALCAPA syndrome, is a rare congenital cardiac disease that can cause myocardial infarction, heart failure and even death in paediatric patients. Only few untreated patients survive until adult age. Here we present the case of a 33-year-old female patient with paroxysmal tachycardia, syncope and little effort dyspnoea. She was diagnosed with ALCAPA syndrome and underwent surgical correction with an alternative technique of left main coronary artery extension to the aorta.",
"keywords": [
"Coronary vessel anomalies",
"ALCAPA syndrome",
"Bland-White-Garland syndrome",
"Adults with ALCAPA",
"Coronary extension technique"
],
"content": "Introduction\n\nALCAPA syndrome, also known as Bland-White-Garland Syndrome, is a rare congenital heart disease, affecting approximately 300,000 newborns in the USA. This disease has 90% of mortality within the first year of life in untreated patients, due to myocardial ischemia and heart failure. Approximately 18-25% of patients with this congenital heart disease reach adulthood, presenting arrhythmias, heart failure and myocardial ischemia1. Treatment of the anomalous origin of the left coronary artery from the pulmonary artery includes several surgical techniques, however they are all associated with important morbidity (21%)2–5.\n\nIn the adult population, surgical correction is more difficult to resolve, due to the heart dimensions and compensatory disorders in coronary circulation to the left ventricle. We present the case of a 33-year-old female with ALCAPA syndrome and mitral valve severe regurgitation, who underwent successful correction with a physiological and anatomical technique.\n\n\nClinical case\n\nA 33-year-old female with medical history of recurrent respiratory infections since childhood, paroxismal tachicardia in adolescence and some syncope episodes in adulthood accompanied by retrosternal pain during exercise. Physical examination revealed a mitral murmur III/IV. The paraclinical diagnostic methods showed anomalous emergency of left main coronary artery from the pulmonary artery, the right coronary dilated, the left ventricle dilated and regurgitant flow in mitral valve (Figure 1).\n\nA – Arrow – RCA dilated arising from the aorta. B – Arrow – LMA arising from lateral aspect of the MPA.\n\n\nSurgical technique and postoperative follow-up\n\nSternotomy and surgical procedure were performed with circulatory support to hypothermia (28°C). The mitral valve was replaced by Mechanical Sorin Carbomedics® valve No 27 to correct valvular dysplasia. The left main coronary artery button was dissected and then connected to a duct constructed with pulmonary wall and bovine pericardium to be anastomosed to the aorta artery. The pulmonary artery was reconstructed with Woven Dacron graft, leaving the previously constructed duct in the back of the Dacron graft. The surgical findings were: right coronary (RCA) dilated and collateral circulation from RCA to left ventricular circulation, LMA arising from the MPA, dysplasia of posterior mitral valve. At 6 months follow-up, the patient remained in functional class I of New York Heart Association and AngioCAT showed patency of the new ductus (Figure 2).\n\n\nDiscussion\n\nALCAPA congenital anomaly is a rare disease that must be surgically treated in the first year of life. However, between 10–15% of patients reach adulthood and clinically manifest rhythm disorders usually attributed to alterations of the cardiac electrical system, which obscures the underlying pathophysiology of myocardial ischemia1–5.\n\nThe blood flow restauration in left main coronary artery from the aorta is the primary objective in the surgical correction of ALCAPA, and there are several surgical options in the paediatric population. Derivation of the left subclavian artery and implementation of an aorto-coronary bypass with saphenous vein or the left internal thoracic artery to the left anterior descending coronary have shown low short-term effectiveness (60%) and high morbidity with stenosis and thrombosis of bypass graft10–14. The Takeuchi procedure is the most used in the paediatric population, however it has a high incidence (> 21%) of supravalvular stenosis of pulmonary artery15.\n\nIn the first year of life, great arteries are not fully developed and tissues are more “flexible”, which allows a coronary reimplantation. However, child’s growth promotes stenosis in short and medium-term7–9. The major anatomical distances and the less “flexible” tissues in adult patients make the surgical restauration of the left main coronary artery blood flow more difficult. Our surgical team resolved this situation with a duct constructed with pulmonary artery wall (80%) and bovine pericardial patch (20%), leaving this duct in anatomical position behind the Woven dacron graft used for restitution of blood flow in main pulmonary artery (Figure 3). We believe that anatomical position of the new duct permits a physiologic blood flow like in a normal heart. In our case, ischemic symptoms resolved and the patient maintained good functional class at 6 months follow-up and full patency of the graft in AngioCAT.\n\nA – LMA taken from the MPA and reconstructed as a tubular structure with bovine pericardium. B – LMA anastomosis to the Ao as in a normal position, MPA reconstructed with a pericardial patch. C – MPA reconstructed with a Dacron graft.\n\nThe ALCAPA physiopathology consists of a relative coronary steal, which promotes low oxigenation in the left myocardial tissue as a consequence of blood flow from pulmonary artery which leads to myocardial ischemia and acute myocardial infarction. The low oxygenation circumstance promotes collateral vessels development and right coronary dilatation, as can be seen in Figure 4. On the other hand, the chronic myocardial ischemia produces papillary muscle and ventricular lateral wall dysfunction, which causes mitral insufficiency. All of this would explain the symptoms presented by the patient.\n\nMitral insufficiency treatment is still under discussion; some authors prefer valvular reconstruction, considering that the failure is due to papillary muscle dysfunction; nevertheless, an important proportion of insufficiency recurrence still exists. In a sense other authors prefer to replace the mitral valve with a valvular prosthesis. In the case that we presented, the surgical team observed a valve dysplasia which prevented valvular reconstruction, so it was decided to replace the mitral valve with a mechanical prosthesis.\n\nTo summarise, the ALCAPA or Bland-White-Garland syndrome treatment is a real surgical challenge in the adult population. However, we believe that the alternative procedure presented in this article consisting of pulmonary artery wall and bovine pericardial construction of a new duct, which connects the left main coronary artery re-establishing a normal anatomical situation and permitting a physiological blood flow to left ventricle, are a viable and probably successful surgical alternative in adult patients without risk of pulmonary stenosis.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details was obtained from the patient.",
"appendix": "Author contributions\n\n\n\nVilá Mollinedo Luis Gustavo (Research, article description and development), Andrés Jaime Uribe (Master Supervision), Jose Luis Aceves Chimal (Article development, Master supervision), Rest (Research, obtaining consent, photography). All authors agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSafaa AM, Du LL, Batra R: A rare case of adult type ALCAPA syndrome: presentation, diagnosis and management. Heart Lung Circ. 2013; 22(6): 444–446. PubMed Abstract | Publisher Full Text\n\nAlsoufi B, Sallehuddin A, Bulbul Z, et al.: Surgical strategy to establish a dual-coronary system for the management of anomalous left coronary artery origin from the pulmonary artery. Ann Thorac Surg. 2008; 86(1): 170–176. PubMed Abstract | Publisher Full Text\n\nGuo HW, Xu JP, Song YH, et al.: Repair of anomalous origin of left coronary artery from the pulmonary artery. Asian Cardiovasc Thorac Ann. 2007; 15(3): 240–242. PubMed Abstract | Publisher Full Text\n\nMohanty SR, Murthy KS, Varghese R, et al.: Evolution of surgical strategies for anomalous left coronary artery. Asian Cardiovasc Thorac Ann. 2001; 9(4): 269–274. Publisher Full Text\n\nMurala JS, Sankar MN, Agarwal R, et al.: Anomalous origin of left coronary artery from pulmonary artery in adults. Asian Cardiovasc Thorac Ann. 2006; 14(1): 38–42. PubMed Abstract | Publisher Full Text\n\nAnil Kumar D, Narasinga Rao P, Kumar RN, et al.: Anomalous left coronary artery: modified direct aortic implantation. Asian Cardiovasc Thorac Ann. 2003; 11(1): 87–89. PubMed Abstract | Publisher Full Text\n\nOhtaki A, Morishita Y, Ishikawa S, et al.: [The Takeuchi procedure for anomalous origin of left coronary artery from pulmonary artery--report of an adult case]. Nihon Kyobu Geka Gakkai Zasshi. 1994; 42(7): 1077–81. PubMed Abstract\n\nGinde S, Earing MG, Bartz PJ, et al.: Late complications after Takeuchi repair of anomalous left coronary artery from the pulmonary artery: case series and review of literature. Pediatr Cardiol. 2012; 33(7): 1115–23. PubMed Abstract | Publisher Full Text\n\nTakeuchi S, Imamura H, Katsumoto K, et al.: New surgical method for repair of anomalous left coronary artery from pulmonary artery. J Thorac Cardiovasc Surg. 1979; 78(1): 7–11. PubMed Abstract\n\nJiménez-Navarro MF, Alegre-Bayo N, Algarra-García J: Diagnosis of ALCAPA syndrome in adults. Rev Esp Cardiol. 2009; 62(10): 1179. PubMed Abstract | Publisher Full Text\n\nOno M, Goerler H, Boethig D, et al.: Surgical repair of anomalous origin of the left coronary artery arising from the left pulmonary artery. Ann Thorac Surg. 2009; 88(1): 275–276. PubMed Abstract | Publisher Full Text\n\nAmanullah MM, Hamilton JR, Hasan A: Anomalous left coronary artery from the pulmonary artery: creating an autogenous arterial conduit for aortic implantation. Eur J Cardiothorac Surg. 2001; 20(4): 853–855. PubMed Abstract | Publisher Full Text\n\nBrown JW, Ruzmetov M, Parent JJ, et al.: Does the degree of preoperative mitral regurgitation predict survival or the need for mitral valve repair or replacement in patients with anomalous origin of the left coronary artery from the pulmonary artery? J Thorac Cardiovasc Surg. 2008; 136(3): 743–748. PubMed Abstract | Publisher Full Text\n\nHirota M, Kawada M, Ishino K, et al.: Anomalous left coronary artery from non-facing pulmonary sinus. Asian Cardiovasc Thorac Ann. 2008; 16(4): 324–326. PubMed Abstract | Publisher Full Text\n\nGinde S, Earing MG, Bartz PJ, et al.: Late complications after Takeuchi repair of anomalous left coronary artery from the pulmonary artery: case series and review of literature. Pediatr Cardiol. 2012; 33(7): 1115–23. PubMed Abstract | Publisher Full Text\n\nLugones I, Kreutzer C, Román MI, et al.: Origen anómalo de la coronaria izquierda en la arteria pulmonar: resultados de la cirugía correctora. Rev Argent Cardiol. 2010; 78(5): 411–6. Reference Source"
}
|
[
{
"id": "15200",
"date": "25 Jul 2016",
"name": "Aiden Abidov",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nA few minor things:\nAbstract: I recommend using a term \"mild exertional dyspnea\" instead of \" little ...dyspnea\".\n\nFigure 3: I would rather remove the Legend from the Figure (makes it less crowded).\n\nFigure 1 - would show a 3rd plate with a longitudinal view of the LM artery (if available) for a better understanding of the anatomy. Just an origin view is not that impressive.",
"responses": [
{
"c_id": "2113",
"date": "03 Aug 2016",
"name": "Luis Gustavo Vilá Mollinedo",
"role": "Author Response",
"response": "First of all, thank you for your evaluation, really helpful, I changed the term of dyspnea for the one you suggested. As for the Figures, we have removed the legend, and as for the CT scan I do not have the one with the longitudinal view. Thanks for your help again."
}
]
},
{
"id": "14947",
"date": "26 Jul 2016",
"name": "Stephan Achenbach",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere is nothing wrong with the description of a surgical technique for ALCAPA repair in the form of a case report. This, however is no proof of superiority over other approaches.\nThe main concern is presentation:\n“ALCAPA syndrome, also known as Bland-White-Garland Syndrome, is a rare congenital heart disease, affecting approximately 300,000 newborns in the USA.” Affecting 300,000 newborns? In which time period?\n\n“This disease has 90% of mortality\"; “important morbidity“; ”surgical correction is more difficult to resolve”; “The paraclinical diagnostic methods showed anomalous emergency of left main coronary artery from the pulmonary artery”; “physiopathology” - examples of incorrect use of the English language.",
"responses": [
{
"c_id": "2114",
"date": "03 Aug 2016",
"name": "Luis Gustavo Vilá Mollinedo",
"role": "Author Response",
"response": "First of all, thank you for your time and recommendations. Addressing all this: We're presenting this approach as an alternative, also we think it's a bit superior having a coronary extension than a baffle procedure, due to the kinking or obstruction probability of these, but also just displaying what we did. I have changed the English use of all the terms and sentences you have proposed, although our English isn't great, we hope to have improved the readability. Finally it was our mistake, the sentence about the incidence should read \"Affecting 1 of 300,000 newborns\". Thanks for your support, any other outlining would be appreciated."
}
]
},
{
"id": "15160",
"date": "08 Aug 2016",
"name": "Roxanne E. Kirsch",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for this concise description of the adult presentation of ALCAPA and offer of an alternate surgical procedure for restoration of coronary blood flow from aorta to LCA.\nThis presents another surgical option for the armamentarium. While you make biologically plausible suggestions as to why this procedure may prove superior, the actual superiority cannot yet be known as only the passage of time and experience with the technique will determine that. This should be more explicitly stated, so as not to be misleading. In that bent, I might revise the final sentence to: \"...permitting a physiological blood flow to the left ventricle is viable and may prove to diminish the risk of pulmonary stenosis post repair of ALCAPA compared with other techniques.\"\nAs a non-surgeon, and a pediatrician, I cannot comment on the likelihood of improved coronary patency personally.\nAlso, it might be helpful for the reader to know the degree of residual mitral regurgitation at the six month follow up, and if you intervened on the valve during the surgical repair (I do not believe you did from my review of this case presentation), and whether this patient's presenting symptoms had abated at the time of 6 month follow up.\n\nMy additional minor suggestions are:\nthe sentence: ...the paraclinical diagnostic methods showed anomalous emergency of left main coronary artery.... should read: \"emergence of the left main coronary artery\"\n\nI would consider an alternate term for the use of \"duct\". This initially seemed like it was referring to a ductus arteriosus before I rapidly realized this was referring to the \"tube graft\" you had created.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1680
|
https://f1000research.com/articles/5-1905/v1
|
02 Aug 16
|
{
"type": "Software Tool Article",
"title": "The Biomedical Research Infrastructure Software as a Service Kit (BRISSKit): technical description",
"authors": [
"Oliver W. Butters",
"Shajid Issa",
"Jeff Lusted",
"Malcolm Newbury",
"Russ Parsloe",
"Nick Holden",
"Robert C. Free",
"Tim Beck",
"Rebecca C. Wilson",
"Paul R. Burton",
"Jonathan A. Tedds",
"Shajid Issa",
"Jeff Lusted",
"Malcolm Newbury",
"Russ Parsloe",
"Nick Holden",
"Robert C. Free",
"Tim Beck",
"Rebecca C. Wilson",
"Paul R. Burton",
"Jonathan A. Tedds"
],
"abstract": "With biomedical research becoming ever more computationally intensive, the challenge is to find sophisticated software tools that can keep pace with new requirements, while still being easy to use and secure. We describe a technical implementation of an infrastructure to manage the full research ecosystem from participant management, to data and sample collection, and finally to data storage, interrogation and analysis. This infrastructure, known as the Biomedical Research Infrastructure Software as a Service Kit (BRISSKit http://www.brisskit.le.ac.uk), is built on open source solutions throughout, and demonstrates that it is possible for a biomedical research platform to be supplied as a service.",
"keywords": [
"CRM",
"LIMS",
"Data collection",
"Data-warehouse",
"Integration Software-as-a-service (SaaS)"
],
"content": "Introduction\n\nThe nature of modern research is to collect ever larger and ever more complex data sets in order to address present day scientific problems, which in turn requires more sophisticated data management1. This increase in size and complexity is particularly apparent in the biomedical research domain, where software tools are having to be rapidly developed to meet these data challenges. This software development is often driven by large research groups who have the resource and expertise to meet their needs. This inevitably results in highly customised software solutions and infrastructure that may not be reusable elsewhere.\n\nSmaller research groups often do not have the resources or expertise to do the equivalent software development themselves. They are then left with no other option than to buy off the shelf (often proprietary) tools in order to meet their needs, or to use tools not designed to do research. Proprietary software tools are often expensive, and often do not allow any user customisation, hindering further reuse. This can then lead to research groups being charged further to have the software modified to meet their requirements.\n\nAn increasingly viable option is to use open source software to build the required research platform. This is an approach that is being actively pushed at ever higher levels - the UK government actively encourages the use of open source software, and have had policies mandating its use when there is no significant cost difference due to its added flexibility (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/78959/All_About_Open_Source_v2_0.pdf).\n\nIncreasingly applications are moving away from being installed locally on client machines (e.g. desktops, laptops etc.) and are being accessed via web browsers e.g. email, word processing, file storage, etc. This has the effect of making them easier to maintain since there is a central install of the software, instead of many local installs on a variety of hardware/operating system/software combinations. It also means that data is not stored locally, thus reducing the risk of data loss or disclosure by the client machine (although there are other risks associated with centralised systems). This approach is often referred to as software as a service (SaaS).\n\nHere we outline the technical aspect of the Biomedical Research Infrastructure Software as a Service Kit (BRISSKit) (http://www.brisskit.le.ac.uk/) project that builds on the trend for open-source and online applications. A subsequent paper (Jonathan A. Tedds, Neil Beagrie, Shajid Issa, Oliver W. Butters, Josh Vande Hey, Scott Wilson, Rebecca C. Wilson, Rowan Wilson, Andrew Charlesworth, and Paul R. Burton - Unpublished report, 2016) will describe use case implementations, the underlying business case, sustainability options, service vision for the platform and proposed further developments and applications.\n\nThere are various web based and open source applications being used (and developed) in biomedical research, for example: the Galaxy project (http://galaxyproject.org) which focusses on genetics analysis, Harvest (http://harvest.research.chop.edu)2 which is a biomedical data discovery framework, ARIES Explorer (http://www.ariesepigenomics.org.uk) which is an epigenetic browser, and tranSMART (http://transmartfoundation.org)3 which is a translational biomedical research knowledge management platform.\n\nThere have been comparable projects in other disciplines, the Virtual Observatory (http://ivoa.net), as developed in the AstroGrid project, e.g. 4 in particular, has influenced the development of this project.\n\nHere we focus on four different open source applications that are applicable to biomedical research studies: CiviCRM, OpenSpecimen, Onyx and i2b2. These four applications were chosen as they formed the core part of the National Institute for Health Research Leicester Cardiovascular Biomedical Research Unit (NIHR-LCBRU) informatics platform, which has been used to recruit thousands of participants into research studies in the East Midlands in the UK (http://www2.le.ac.uk/research/current-research/bru/our-research/research-facilities/informatics-platform).\n\nCiviCRM. CiviCRM (https://civicrm.org) is a web-based open source (GNU AGPL v3) constituent relationship management tool built on top of Drupal (https://drupal.org). It is designed to manage the contact details of individuals and their relationships with things. It also manages the means to contact individuals. It can be configured to have almost any number and type of data field for each individual added to it. This configurability makes it an ideal tool to track study participant details, e.g. names, addresses, phone numbers etc. CiviCRM then adds the ability to model relationships between participants e.g. mother-child, doctor-patient, work colleague etc. CiviCRM also adds the concept of organisations, these can be used to model e.g. a household which a number of participants could belong to, a hospital that participants are patients in etc. Furthermore, CiviCRM adds extra value in its case functionality - this allows a series of activities to be defined which may model individual stages of a biomedical study, for example: fill in a consent form, take a blood sample, schedule an appointment etc. These activities may be linked together sequentially, or in a more complex non-sequential way with additional conditional logic. This mirrors the work flow that a biomedical research study may have.\n\nGiven these features it is easy to see how CiviCRM could serve as the main study management application in a biomedical study.\n\nCiviCRM is written in PHP, and uses a MySQL database. The source code is available at: http://sourceforge.net/projects/civicrm\n\nOpenSpecimen. OpenSpecimen (formally known as CaTissue) (http://www.openspecimen.org/) is an open source (BSD 3-clause licence) web-based biobanking management system. It was originally developed as part of the U.S. National Cancer Institute’s caBIG program as CaTissue. CaTissue was then forked by a commercial company and re-branded as OpenSpecimen (still keeping it open source). They now maintain the core code base, and offer support and hosting. It has a highly configurable object model, making it possible to model almost any type of storage infrastructure used in biobanking (e.g. boxes on shelves, in freezers, in rooms, in buildings etc.). Samples can then be put into the system and tracked as e.g. containers are moved or samples are checked out etc.\n\nOpenSpecimen is written in Java and uses a MySQL database. The source code is available at: https://github.com/krishagni/openspecimen\n\nOnyx. Onyx is an open source (GNU GPL v3) web-based data collection tool developed by the Canadian company OBiBa (http://obiba.org). It is primarily used to collect questionnaire data from study participants, and was developed with the aim of collecting data from over 300,000 participants in the Canadian Partnership for Tomorrow program. Onyx is written in Java and uses a MySQL database. It is available at https://github.com/obiba/onyx\n\ni2b2. i2b2 (Informatics for integrating biology and the bedside) (https://www.i2b2.org) is an open source (custom written licence - https://www.i2b2.org/software/i2b2_license.html) data warehouse framework built by Partners Healthcare System5. At its core are several ‘cells’, each providing a specific piece of functionality e.g. identity management, ontology management, data storage, natural language processing, web client etc. These cells are arranged together into the i2b2 ‘hive’, in which each cell can communicate with the others via XML based web services. This allows biomedical data from multiple sources to be stored with ontological codes and presented side by side. The end users can then query the multi-source data using a web browser. Using this functionality some analysis can be done on the data, or new cohorts of participants generated based on some phenotypic criteria.\n\ni2b2 is written in Java and uses a Microsoft SQL server, an Oracle or a postgreSQL database. It is available at https://www.i2b2.org/software/\n\nEven with the suitability and availability of the applications above, they have not been widely adopted in the field. We think this is due to the presence of three main barriers to widespread adoption: installation and maintenance, hosting and integration.\n\nInstallation and maintenance. Each of the user facing applications are freely available for anyone to download and install without any charge or mandatory contracts. They are all provided with some installation instructions which will guide a user through installing and configuring the applications. However, a high degree of technical expertise is still usually required to perform these steps.\n\nOnce the applications have been installed they need to be maintained, this will usually involve upgrading the software when new versions are released, troubleshooting any problems with the software, and regularly backing up the data. Again, a high degree of technical expertise is required for this.\n\nHosting. An important decision to take when planning on running one of these user applications is where it should be installed. Since each application is accessed through a web browser it is vital that consideration is given to accessibility and security. A key factor in the choice of host has to be the physical location of data centres, since each data centre is subject to the local laws of the country in which it is physically based. At a more local level, hosting providers do not all offer the same service, some have high levels of security standards which they evidence with certifications like ISO 27001, others do not. Some have a direct connection to the UK academic network JANET (http://ja.net), others to the UK NHS network. Wherever sensitive patient data is involved a careful and thorough approach to information governance is essential. In the UK compliance with the NHS Information Governance Toolkit (https://www.igt.hscic.gov.uk/) is often required.\n\nIntegration. Each application is a valuable resource in its own right, but even greater value can be achieved when they are integrated together. A simple example of this would be where multiple data collection tools are able to automatically export their data and have it imported into a central data warehouse. This would allow data from multiple sources to be analysed at once.\n\nThis integration process is perhaps the most difficult to overcome of the barriers to widespread adoption, since it requires detailed technical knowledge of multiple systems.\n\nWhile the three main barriers to widespread adoption (the technical know-how to install and maintain the applications, the facility to host the applications in a secure environment, and the facility to integrate the applications together and to external applications) have been individually overcome, to a greater or lesser extent by various groups, a significant amount of development and time is generally needed.\n\nIt is with this backdrop that the BRISSKit project was conceived and exists - it aimed to provide access to a suite of mature open source applications, hosted in a secure environment, integrated together and accessed via a web browser. The intended end-user base for BRISSKit was that of groups with multiple users who may or may not be co-located, and who do not necessarily have the technical experience (or resources) to set up and maintain the software themselves.\n\nThe user facing applications chosen to achieve this are those outlined earlier i.e. CiviCRM (v4.1), OpenSpecimen (v1.2-plus2.0), Onyx (v1.9) and i2b2 (v1.5).\n\nAlthough BRISSKit has been primarily focused at biomedical research groups, these tools can be adopted in a similar way by other disciplines also.\n\n\nMethods\n\nThis section describes how BRISSKit addressed the three main barriers to widespread adoption outlined earlier. The core infrastructure design subsection covers installation, maintenance and hosting, and the subsequent section addresses integration.\n\nBelow is an outline of the main design choices and components of the core infrastructure, on which the client facing software is installed. It begins at the bottom of the stack with the virtualisation/operating system layer, it then moves up a layer and addresses the configuration of the operating system, then up to the actual install of the software, finally moving to the overarching monitoring.\n\nVirtualisation layer. A key objective for the infrastructural design is to make the platform easily accessible, deployable and scalable. A cloud based environment facilitates these needs and allows the rapid provision of new resources as needed.\n\nAn early development decision was to use an Infrastructure as a Service (IaaS) provider. This allowed the maximum degree of customisation when designing and running the platform. With IaaS, virtual machines (VMs) can be provisioned with the required specifications as needed. In order to take into account the issues around the physical location of data centres, a UK based hosting provider with UK based data centres was used - Eduserv’s cloud compute solution (http://www.eduserv.org.uk). Eduserv also has a direct connection to JANET, allowing fast transfer of data to/from UK universities. Eduserv provided VMWare’s vCloud Director interface (http://www.vmware.com/products/vcloud-director/) giving a software defined data centre.\n\nOne of the features missing from the Eduserv offering was direct NHS connectivity. In order to meet this need we worked closely with the University Hospitals Leicester Trust to develop the BRISSKit platform to run on their internal (N3 connected) infrastructure, which ran VMWare’s vSphere (v5).\n\nWhile both of the above infrastructures use proprietary VMWare software, none of the proprietary methods (e.g. for the provisioning of new VMs) were used i.e. the BRISSKit platform could be run on any virtualisation technology or provider (e.g. Amazon, Azure etc.).\n\nEncapsulation. When designing a cloud based software solution it is important to consider how different users may interact with one another. Clearly each research group wants their data to be completely demarcated from any other research group. With this in mind the platform was designed to completely encapsulate one instance of the software stack, not allowing any communication with any other instances of the stack. Initially this was achieved using VMWare’s vApp functionality, with each application having its own VM. However, this was deemed to be too closely tied to one vendor’s proprietary methods. Later this was moved to a software defined vApp (still with a separate VM for each application). Through an automated method the VMs were grouped together and isolated by the firewall, so could only communicate with other VMs in the same vApp. This allowed a way of building an instance of the infrastructure in a way that is completely agnostic to the underlying virtualization technology. This can be visualised in figure 1 where two instances of the software stack are shown running side by side, but are isolated from one another.\n\nTwo independent research groups are shown (group A and group B) illustrating the encapsulation of resources. The lines show the only network routes into and out of the infrastructure.\n\nConfiguration management. A key design choice for the management of the VMs was to use the Puppet configuration manager (v2.7) (https://puppetlabs.com/) to manage all software installation, configuration, user access etc. in the VMs, up to the point where the user applications can be installed. This allows the configuration to be managed in a declarative way, and information about VMs to be collected into a central resource. There are several benefits to this approach; the declarative nature means that the client VMs end up being configured in the required way- the process taken to get there is not important. Development of scripts which handle the configuration in a specific order is therefore not necessary. It also makes the type of guest operating system (OS) on the VM less important since e.g. software listed as being required is installed by the Puppet client, regardless of the different software management packages each OS has.\n\nAnother advantage of all configuration being managed centrally is that there is no real need to log onto specific client VMs to make changes such as new firewall rules etc. One final major advantage of using Puppet is the central gathering of information about the clients. Puppet has, at its core, a central database which logs information about the state of the clients, this can include server info such as uptime etc., as well as versions of software installed, which is a useful tool for auditing.\n\nAll of the configuration is put together in a central catalogue that each VM can query. Based on the role of the VM the catalogue items are automatically configured to e.g. set up the appropriate firewall rules etc. The overarching catalogue items developed as part of this project are listed in table 1.\n\nInstantiation. Encapsulation is achieved based on the name and role the VM is given when it is created. As an example, imagine a research group using BRISSKit called groupA, a VM called groupA-civicrm would be created. Puppet looks at this name and deduces that it belongs to the software defined vApp for groupA, and it’s role is as a CiviCRM VM. Puppet then takes this new VM’s IP address and allows access through the firewalls on each of the other VMs in this vApp to this new VM, and adds a hostname entry to each. The other VMs in this vApp will then be able to connect to groupA-civicrm as required. At the same time the reverse proxy has an entry added to its rules so any web traffic meant for CiviCRM is directed to the correct VM. It also automatically adds an entry to the central nagios monitoring server. Once this generic configuration management has finished, Puppet then applies the catalogue entry relevant for this role (in this case - installing Apache, PHP libraries and the MySQL client). The end point of this process is a VM which is ready to have the client software (CiviCRM in this case) installed on it.\n\nWeb access. As mentioned previously, all of the user applications are primarily accessed via a web interface. To tie these together a reverse proxy was implemented, this allowed a subdomain to be defined per research group and each application to sit below that, e.g.\n\n• groupA.brisskit.le.ac.uk/civicrm\n\n• groupA.brisskit.le.ac.uk/openspecimen\n\n• groupA.brisskit.le.ac.uk/i2b2\n\n• groupB.brisskit.le.ac.uk/civicrm\n\n• etc.\n\nThe open source reverse proxy software Pound (v2.5) (www.apsis.ch/pound) was used to achieve this.\n\nAll web traffic was encrypted from the client to the reverse proxy with SSL certificates.\n\nClient software installation and maintenance. A significant amount of effort was devoted to automating the install of each user facing application. This gave the benefit of being able to deploy an application very quickly and in a standard way, thus avoiding any mistakes that may occur due to human error. In order to do this a common platform was defined in which to start the installation from. This consisted of an 64bit Ubuntu 12.04 Long Term Support operating system (http://www.ubuntu.com) with all configuration centrally managed with Puppet. This meant that the installation process for each application consisted of deploying a new instance of Ubuntu, the Puppet master would then automatically configure it ready to run the application installation.\n\nThe application installation scripts were all managed in version control, and then built with the software build tool Maven (v2.2) (http://maven.apache.org). Periodically, copies of the Maven built artefacts were deposited in our local Nexus repository, along with all the relevant dependencies, or links to remote repositories (Nexus software: http://www.sonatype.com/nexus-repository-oss). This process gave a standard installation procedure that could be followed for each of the applications, despite them requiring very different processes and dependencies in their native form.\n\nExtra guidance and documentation on the install of the core applications that has been built up over the project is also available on the project website at http://www.brisskit.le.ac.uk.\n\nReporting. An essential part of any infrastructure is monitoring of servers and applications. Within BRISSKit, Nagios (v3) was implemented (https://www.nagios.org/). Nagios is an open source (GNU GPL v2) monitoring solution that follows the client-server model. All of the VMs report in periodically on their status to the central Nagios server. Different VMs report different measures based on their role. There are a core set of measures (CPU load, disk usage etc.) that all report, but on top of this there are others - the MySQL VM reports the status of its MySQL server for instance.\n\nThis set up facilitated the proactive monitoring of the infrastructure and fixing of problems as they happened, moreover, developing problems could be fixed before they manifested. This also served as a means of measuring resource usage and therefore facilitating cost effective use of compute resource.\n\nThe distributed nature of the BRISSKit infrastructure meant that the Nagios server could not always instigate active checks. Passive checks were therefore run across the infrastructure, instigated from the clients. This was achieved by using the NRDP Nagios plug-in (https://exchange.nagios.org/directory/Addons/Passive-Checks/NRDP–2D-Nagios-Remote-Data-Processor/details) on the clients along with scheduled cron jobs. All of which was managed with Puppet.\n\nUse case. In order to describe the integration model developed, an end to end use case needs to be outlined first. Assume a study is using all four applications, the study behaviour can be defined in CiviCRM, into which study participants can be added. The study definition then means that these participants could be passed to the other applications in the stack. Onyx would then be ready for a participant to fill in a questionnaire as it would have e.g. a name pre-populated and an appointment time specified. OpenSpecimen would also be ready for a sample to be input. Once the data is collected it is automatically imported into i2b2, and data from the different sources (i.e. Onxy and OpenSpecimen) about the same individual is joined together. It is here that analysis of the data would happen. If there was a new cohort that emerged from the analysis, then they could be imported back into CiviCRM to be re-identified, and then invited back for more tests.\n\nIt is with this use case in mind that we describe the integration.\n\nThe layers. The BRISSKit architecture can be thought of as different layers, each containing a different category of application. Each layer can communicate with the other layers in a well defined way via the internal application programming interface (API). In this way the architecture can be split into three distinct layers based on the category of application it contains - management, data collection and data warehousing. This allows the user facing applications to be categorised into one of those three; CiviCRM - management, OpenSpecimen and Onyx - data collection, and i2b2 - data warehousing. The layers are stacked, with the management layer at the top, the data collection layer in the middle, and the data warehousing layer at the bottom - see figure 2. Messages are sent between the layers, through the internal API.\n\nThis layered architecture also serves as a means to categorise user access. For example, administrative staff may have access to identifiable contact details in CiviCRM, lab and data collection staff may have access to OpenSpecimen and Onyx and researchers access to non-identifiable data in i2b2. In this way data is very clearly segregated based on user roles.\n\nThe internal API. The layered architecture allows us to define standard messages that get sent between the layers. Each individual application must then adhere to the standard message definition when it sends messages between layers. Figure 2 illustrates this.\n\nBy adopting this layered structure it becomes relatively straightforward to add new components to the software stack - they just need to be able to communicate with the adjacent layers. This can be achieved by wrapping each new application up in such a way that it only communicates via the wrapper (the orange part in figure 2).\n\nIn order to make the API easy to use it is being developed as a REST interface. This also allows the API to be portable and scalable. Table 2 lists some pseudo-API calls to illustrate the layers.\n\nThe layers correspond to data going from the management layer (M) to the data collection layer (C), and to the warehousing layer (W).\n\nThe approach taken to implement the internal API will be different for each application due to the different technologies used in each.\n\nCiviCRM has a mature REST API allowing interaction with it’s core functions. In order to achieve the BRISSKit internal API calls the native CiviCRM API was wrapped in our own functions. The add_participant_to_(object) function was triggered when participants were added to a study in CiviCRM and had an appointment booked. This passed the relevant participant information to the data collection tools.\n\nOpenSpecimen did not have a mature API available during the development of this phase of the project (it does now). In order to fit it into our infrastructure we therefore had to call core Java classes to manipulate the data. The developed integrations accepted a participant from CiviCRM and created the sample collection stub, then once the samples were taken the relevant information extracted and pushed to i2b2. Now the API is more mature (and RESTful) these calls could be migrated to use it.\n\nOnyx does not have an API available. It does however have routines that can be called to load and export data. These were wrapped into BRISSKit functions to allow participants to be added, and for the data to be extracted and pushed to i2b2.\n\ni2b2 does not have an API as such, it’s modular design uses internal API calls to facilitate communication between cells, but this is not intended for external use. Modifying the Clinical Research Chart (CRC) loader facilitated loading data from the collection tools (import_data). We also developed a package so that groups of participants defined in i2b2 could be pushed back to CiviCRM. This enabled a work flow where a researcher could define a group of participants in i2b2 based on questionnaire answers and availability of samples without having access to their contact details, push that group back to CiviCRM to be re-identified (reidentify_cohort), and then followed up for more tests.\n\nID numbers and linking data. CiviCRM generates a unique random ID for each participant. This is the pseudonymised ID pushed to each data collection application in the layer below. Once the data is exported from each data collection application it is linked together at the data warehousing layer using this ID.\n\nOntology builder. Ontologies are an important aspect of the BRISSKit data warehousing layer since they form an integral part of i2b2’s functionality, and give the data meaning. Part of the integration of new applications in the stack therefore requires i2b2 to have an ontology describing the data in order to be able to understand it. Within the infrastructure, OpenSpecimen and Onyx automatically generate their own ontologies based on the data structures in the applications. This gets passed to i2b2 so the data can be queried.\n\ni2b2 does not specify which ontology it has to use, so the automatically generated ones are considered ‘nominal ontologies’, i.e. they do not necessarily conform to one of the standard ontologies such as e.g. SNOMED CT.\n\nSince it is possible that not all data will come from the core applications (i.e. CiviCRM, Onyx and OpenSpecimen) - some may need to be imported into i2b2 from external sources - an ontology building tool was developed. This plugged into the National Center for Biomedical Ontology’s BioPortal service (http://bioportal.bioontology.org/) and allowed ontology codes from standard ontologies to be used.\n\n\nFuture development\n\nGiven the modular nature of the infrastructure design, and the relative ease of integrating new applications, we are planning on adding more applications to the stack. Opal (http://obiba.org/node/63) is a data warehouse built by the same group that developed Onyx. One of the key features of Opal is that it integrates with the analysis tool DataSHIELD6,7, allowing analysis across multiple data warehouses in a non-disclosive way. Early implementations focussed on simultaneous epidemiological analysis across multiple international birth cohorts but this approach could be just as valuable in e.g. enabling analysis of studies across multiple hospital sites in one territory. At the data collection layer we have done some work to integrate the Research Electronic Data Capture tool, REDCap (http://www.project-redcap.org), into the stack. Going forward we will develop this integration so it becomes a core part of the BRISSKit stack, and is available to groups who have the appropriate agreement to use it. The internal API needs to be formalised somewhat, then an external API can be developed to connect to it, thus making it easy to integrate with third-party systems e.g. importing data from external systems. It is likely that we would use openESB (http://www.open-esb.net/) to further extend this functionality.\n\n\nSummary\n\nWe have shown that it is possible to install and integrate a suite of mature open source applications for use by biomedical researchers. Moreover we have demonstrated that these applications can be installed in a cloud environment and isolated in such a way that multiple research groups can share the same infrastructure, but have their data completely separate from one another. We believe this a viable alternative to local installations of proprietary software, and logically leads to the idea of a research platform as a service that could be offered to research groups. A subsequent paper (Jonathan A. Tedds, Neil Beagrie, Shajid Issa, Oliver W. Butters, Josh Vande Hey, Scott Wilson, Rebecca C. Wilson, Rowan Wilson, Andrew Charlesworth, and Paul R. Burton - Unpublished report, 2016) will describe use cases and outline sustainability options for the BRISSKit platform.\n\n\nSoftware availability\n\nLatest source code: https://github.com/brisskit-uol\n\nArchived source code as at time of publication:\n\nCiviCRM http://dx.doi.org/10.5281/zenodo.573848\n\nOnyx install: http://dx.doi.org/10.5281/zenodo.573809\n\nOpenSpecimen install: http://dx.doi.org/10.5281/zenodo.5737710\n\ni2b2 install: http://dx.doi.org/10.5281/zenodo.5737411\n\nPuppet: http://dx.doi.org/10.5281/zenodo.5736512\n\nSoftware license: BSD 3-clause license.",
"appendix": "Author contributions\n\n\n\nOB, SI, JL, RP & RF all contributed to the overall technical development of BRISSKit. TB designed the ontology builder aspect.\n\nJT was PI and conceived the original BRISSKit concept and design building on early use cases developed with NH & colleagues in the NIHR Cardiovascular Biomedical Research Unit. JT managed the project through subsequent stages at the University of Leicester including leading on all strategic, funding, partner, contractor and institutional liaison in particular with Jisc, University Hospitals Leicester NHS Trust and the UK National Institute for Health Research.\n\nMN & NH were heavily involved in the technical design of BRISSKit. PB was centrally involved in strategic development of the BRISSKit project.\n\nOB, JT, RW, SI & PB prepared and reviewed the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe bulk of the development work took place at the University of Leicester (PI J.Tedds) and was supported by Jisc under the HEFCE University Modernisation Fund - Shared services and the cloud programme 2011:2012 with additional funding during 2012:2013 and 2014:2015 from Jisc and the University of Leicester. OB is funded under a strategic award from MRC and Wellcome Trust for the ALSPAC project [102215/Z/13/Z]. The University of Bristol component of the work described in this article is funded as a central element of the research program of the Data to Knowledge (D2K) Research Group, jointly supported by funding from: the European Union’s Seventh Framework Programme BioSHaREEU [261433] (Biobank Standardization and Harmonization for Research Excellence in the European Union) and BBMRI-LPC [313010] (Biobanking and Biomolecular Resources Research Infrastructure - Large Prospective Cohorts). MRC: the Welsh and Scottish Farr Institutes, MRC funded E-Health Informatics Research Centres (EHIRCs) [MR/K006525/1; MR/K007017/1]. Wellcome Trust & MRC: 58FORWARDS [108439/Z/15/Z] (The 1958 Birth Cohort: Fostering new Opportunities for Research via Wider Access to Reliable Data and Samples).\n\n\nAcknowledgements\n\nThe authors acknowledge the contributions of a wide range of academics and support staff at the University of Leicester during this work including in the College of Medicine, Biological Sciences and Psychology, IT Services, Research and Enterprise Division and Library. We also thank the many NHS staff at University Hospitals Leicester NHS Trust including Chris Greengrass, Tim Skelton, John Clarke, Andy Carruthers, David Rose, Richard Bramley, Alison Goodall, Nilesh Samani, David Wynford-Thomas, Kevin Schurer, Anthony Brookes, Mary Visser, Robert Feakes and a range of other partners, collaborators and external contractors involved; early pilot evaluators led by Dr Tito Castillo at Great Ormond Street Hospital, University College London, and at the School of Cancer Studies, University of Birmingham, led by Paul Mason. We highlight the long standing and much valued insight and support from Peter Knight, Deputy Director at the Department of Health; John Milner, Simon Hodson, Rachel Bruce, Daniela Duca, Catherine Grout, Martin Hamilton and colleagues at Jisc; and Neil Beagrie and Daphne Charles at Charles Beagrie Ltd.\n\n\nReferences\n\nWhyte A, Tedds J: Making the case for research data management. DCC Briefing Papers. Edinburgh: Digital Curation Centre; 2011. Reference Source\n\nPennington JW, Ruth B, Italia MJ, et al.: Harvest: an open platform for developing web-based biomedical data discovery and reporting applications. J Am Med Inform Assoc. 2014; 21(2): 379–383. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAthey BD, Braxenthaler M, Haas M, et al.: tranSMART: An Open Source and Community-Driven Informatics and Data Sharing Platform for Clinical and Translational Research. AMIA Jt Summits Transl Sci Proc. 2013; 2013: 6–8. PubMed Abstract | Free Full Text\n\nTedds J: Science with the Virtual Observatory: the AstroGrid VO desktop. \"Multi wavelength astronomy and the Virtual Observatory\" conference, EuroVO-AIDA program, European Space Astronomy Centre, Spain. arXiv: 0906.1535 [astro-ph.IM]. 2009. Reference Source\n\nMurphy SN, Mendis M, Hackett K, et al.: Architecture of the open-source clinical research chart from Informatics for Integrating Biology and the Bedside. AMIA Annu Symp Proc. 2007; 548–552. PubMed Abstract | Free Full Text\n\nWolfson M, Wallace SE, Masca N, et al.: DataSHIELD: resolving a conflict in contemporary bioscience--performing a pooled analysis of individual-level data without sharing the data. Int J Epidemiol. 2010; 39(5): 1372–1382. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGaye A, Marcon Y, Isaeva J, et al.: DataSHIELD: taking the analysis to the data, not the data to the analysis. Int J Epidemiol. 2014; 43(6): 1929–1944. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLusted J, Stevens I, Issa S: civicrm-install v1.0. Zenodo. 2016. Publisher Full Text\n\nLusted J, Issa S: onyx-install-procedures v1.0. Zenodo. 2016. Publisher Full Text\n\nButters O, Issa S: caTissue-v1.2plus2.0-install-procedures v1.0. Zenodo. 2016. Publisher Full Text\n\nButters O, Lusted J, Issa S: i2b2-install-procedures v1.0. Zenodo. 2016. Publisher Full Text\n\nButters O: puppet_setup v1.0. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "19399",
"date": "26 Jan 2017",
"name": "Johan Nyström-Persson",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the BRISSKit, a software infrastructure for biomedical research consisting of several open-source applications. The applications have been virtualised and deployed in a cloud environment, as well as wrapped in layers with custom APIs to allow for inter-application communication. This aims to make use of open source software more viable for researchers, by solving problems of installation, maintenance, hosting and application integration, while also taking into account information security and encapsulation. The authors have also ensured that BRISSKit remains independent of any specific cloud infrastructure vendor, ensuring portability.\nThe authors address an important problem and provide a valuable study on how open source software can successfully be deployed, integrated and used as a service in the biomedical research domain. The scenario being studied is described in considerable detail and has clear practical value, and the source code is publicly available. As such, I believe that the paper and the toolkit is a nice contribution and should be published. However, I have some concerns that the authors might address.\nImproving the ease of access, configuration, integration and use of open source software is a worthy goal. The drudgery of configuration and integration may certainly represent a significant barrier to adoption in many cases, and supplying the software as a service has potential to go a long way towards removing these barriers.\nMany X-as-a service offerings exist, and it would be useful to define more precisely which area the framework targets. For example, what are the essential needs of biomedical, as opposed to other, kinds of research? What are the essential needs of research software and infrastructure as opposed to infrastructure targeting other kinds of users or activities? If this is clearly stated, it becomes easier to evaluate the approach.\nI believe that one essential concern in research software - as opposed to, say, industrial or consumer oriented software - is that method innovation may be a routine part of everyday work, as research goes hand in hand with method and tools development. This method innovation may often include changes or enhancements to software. Thus, while it is important to remove inessential tedious procedures that act as barriers to adoption, it is equally important that customisation, extension and tinkering is possible, even for unsophisticated users, if the framework is to have maximal relevance and impact. Ideally, the framework should provide convenience without erecting any new barriers. Thus, I would like to see a discussion of the amount of effort needed to add new applications to an existing BRISSKit installation, or to make customisations or modifications, from the point of view of a small research group with modest resources and software skills. It is important that this autonomy is not lost if researchers choose to depend on service/cloud-based offerings for their basic infrastructure. Alternatively, if BRISSKit is only intended for research scenarios where the software methodology has been fixed in advance, then this should be made clear.\nOne of the main difficulties that seems to remain in deploying and using multiple software applications in concert, even after the BRISSKit methodology has been adopted, is making applications talk to each other. BRISSKit solves this by dividing the applications into layers and specifying that each layer only has to talk to adjacent layers. It is not clear to me that the layer model is always applicable (unless BRISSKit specifically only targets environments that need the three layers of management, data collection and data warehousing - in which case this should be stated). What would be a good architecture if the three-layer model does not apply in some scenario? There is also a lack of standards or guidelines for the API design, which would mean in the worst case that the integration problem is not solved, only transposed into the problem of API design and development. It would be useful to see some design principles here. In general, I think it would be good to state the scope of the BRISSkit design more precisely and explicitly.\nWhile the paper does describe the framework in detail, its success with respect to the stated objectives is not evaluated - for example, from the point of view of users and researchers. The authors do mention that another publication with use cases is forthcoming, but I believe that this paper would benefit from at least a brief evaluation or theoretical justification of BRISSKit’s success in meeting the stated objectives.",
"responses": []
},
{
"id": "19828",
"date": "01 Feb 2017",
"name": "Christian Ohmann",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSupport of biomedical research with a SaaS toolkit is a relevant issue and the selection of open source tools is the right way. There are, however, some points to be discussed:\n\nThe selection of the different Open Source tools to be integrated in the SaaS should be discussed and motivated.\n\nIt is not clear how the elicitation of formal requirements was handled. Was there an analysis of the research flow? Who are the stakeholders, the users, the data owners, etc.?\n\nThe paper provides a technical description. Nevertheless, without testing/evaluation data and a use case (which will be published in another paper), the exercise is more theoretic. It is not demonstrated how the different solutions interact efficiently with each other in the cloud.\n\nData protection issues for cloud computing are not sufficiently discussed. What happens with personal or pseudonymised data in the cloud? How is informed consent for data (re) use handled with the system?\n\nThere should be a discussion on the pros and cons of the approach taken. The solution developed is partly UK-specific and the question should be raised how this can be applied in other countries and different locations.",
"responses": []
},
{
"id": "17938",
"date": "18 Apr 2017",
"name": "Adam Huffman",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe reasoning behind the use of open source applications with an awareness of the difficulties they can present, and the decision to host on infrastructure as a service platforms is sound, and well presented. However, a more detailed consideration of the individual application choices is required, beyond their previous use in local studies. For example, which alternatives to CiviCRM were considered and why were they rejected?\nThe complications and downsides of using a third-party cloud resource (which would be likely required for someone aiming to install BRISSKit themselves) are not examined in any detail. More explanation of the 'software-defined vApp' approach is required, to enable a similar level of isolation between instances to be achieved on non-VMware platforms.\nIssues of data security are raised, and addressed, but not much guidance is given for those wishing to adopt BRISSKit who do not have access to the specific service provider chosen for the project.\nUbuntu 12.04 reaches end of life in April 2017. No discussion is given of changes that may be needed (for instance in the Puppet modules) to accommodate porting to a newer version of Ubuntu. Moreover, further detail is required to give confidence that the applications can be installed on other operating systems than those mentioned. This might include providing examples of operating system-agnostic Puppet code.\n\nThe specific version of Puppet cited in the report has reached end-of-life, so similar concerns apply there regarding updating to a supported version.\n\nGiven the importance of data security, it is surprising that the mention of SSL certificates does not include any mention of certificate authorities and client browser validation of certificates.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1905
|
https://f1000research.com/articles/5-1904/v1
|
02 Aug 16
|
{
"type": "Opinion Article",
"title": "Syphilis may be a confounding factor, not a causative agent, in syphilitic ALS",
"authors": [
"Bert Tuk"
],
"abstract": "Based upon a review of published clinical observations regarding syphilitic amyotrophic lateral sclerosis (ALS), I hypothesize that syphilis is actually a confounding factor, not a causative factor, in syphilitic ALS. Moreover, I propose that the successful treatment of ALS symptoms in patients with syphilitic ALS using penicillin G and hydrocortisone is an indirect consequence of the treatment regimen and is not due to the treatment of syphilis. Specifically, I propose that the observed effect is due to the various pharmacological activities of penicillin G (e.g., a GABA receptor antagonist) and/or the multifaceted pharmacological activity of hydrocortisone. The notion that syphilis may be a confounding factor in syphilitic ALS is highly relevant, as it suggests that treating ALS patients with penicillin G and hydrocortisone—regardless of whether they present with syphilitic ALS or non-syphilitic ALS—may be effective at treating this rapidly progressive, highly devastating disease.",
"keywords": [
"amyotrophic lateral sclerosis",
"ALS",
"syphilitic ALS",
"GABA antagonist",
"penicillin G",
"benzylpenicillin",
"hydrocortisone",
"circadian rhythm"
],
"content": "Introduction\n\nAmyotrophic lateral sclerosis (ALS, also known as Lou Gehrig’s disease) is a devastating disease with an average post-diagnosis life expectancy of 3–5 years. The clinical manifestations of ALS are well described and include the progressive wasting of muscle mass, reduced muscle coordination, dysphagia, dysarthria, and fatal respiratory depression1–3. Several observations regarding the putative pathogenesis of ALS have been reported1–3; however, although more than 130 years have passed since it was first described, the pathogenesis of ALS remains poorly understood.\n\nOne of the many unresolved mysteries surrounding the pathogenesis of ALS is the presumed association of ALS with syphilis (so-called “syphilitic ALS”). The most puzzling observation in syphilitic ALS is that it is the only form of ALS that has been reported to have been treated effectively; specifically, in 1990 El Alaoui-Faris reported five cases of syphilitic ALS in which the patients were followed for 5–13 years after receiving treatment for syphilis4. Interestingly, these five reports originated from one neurology center in Morocco. In this center, five patients tested positive for syphilis in their serum and/or cerebrospinal fluid (CSF) and were successfully treated using a specific combination of penicillin G and hydrocortisone, the standard of care for neurosyphilis in Morocco at the time. Because their ALS symptoms also resolved during treatment, and because these patients no longer tested positive for syphilis, the group concluded that the patients’ ALS symptoms were caused by syphilis, and they concluded that neurosyphilis can lead to ALS, giving rise to the syphilitic ALS hypothesis4.\n\nHere, I postulate that in countries in which syphilis is highly endemic, syphilis may actually be a confounding factor in syphilitic ALS, and this factor should be considered carefully before drawing conclusions regarding the efficacy of treatment in these populations. Based upon a careful review of the aforementioned five reported cases of syphilitic ALS, I conclude that the only evidence for the existence of syphilitic ALS as a specific disease entity stems from fact that the patients’ ALS symptoms resolved simultaneously with the shift from syphilis-positive to syphilis-negative following treatment with penicillin G and hydrocortisone. Thus, I propose the novel hypothesis that the successful treatment of syphilitic ALS with a specific penicillin G and hydrocortisone regimen is independent of the treatment of syphilis. Specifically, I propose that the anti-syphilis treatment treated these patients’ ALS directly via the off-target pharmacological activity of penicillin G (e.g., as a GABA receptor antagonist) and/or hydrocortisone.\n\n\nCase reports of five patients with syphilitic ALS who were followed for 5–13 years post-treatment\n\nThe five reported cases of treated syphilitic ALS published in 1990 are summarized in Table 1 and Table 24. These cases originated from a neurology center in Morocco that was highly experienced in the diagnosis of ALS4. Out of a total of 40 ALS patients who were diagnosed at this neurology center, five were diagnosed with syphilitic ALS. The age at onset of ALS was 27–48 years, and the clinical symptoms included classic features of ALS, including functional impairment of the upper limbs in three patients and spastic paraparesis in two patients4. Before receiving treatment, progressive neurological dysfunction had been present from several months to up to 3 years. Two patients also displayed symptoms that are not typically present in ALS patients: patient 3 reported sudden regressive deafness, which was later found to be caused by meningogenic labyrinthitis, and patient 5 had horizontal nystagmus.\n\n\nDiscussion and conclusions regarding syphilitic ALS\n\nSyphilitic ALS is an intriguing phenomenon, as it is the only form of ALS ever reported to have been cured4. Moreover, syphilitic ALS has only been reported in Morocco, a country in which syphilis is highly endemic, thereby greatly increasing the likelihood of presenting in patients with ALS. Lastly, neurosyphilis can be a major confounding factor when diagnosing ALS. Therefore, these five case reports of putative syphilitic ALS were reanalyzed in order to explore plausible explanations for these patients’ resolution of ALS symptoms.\n\nFirst, consider the evidence supporting the existence of syphilitic ALS, which stems solely from the observation that ALS symptoms resolved after treatment with penicillin G and hydrocortisone in patients who also tested positive for syphilis. Because a cure for ALS had never been reported, and because penicillin G is an effective treatment for syphilis and neurosyphilis, the authors concluded that these patients had syphilitic ALS. Their reasoning was based on the fact that the treatment regimen—in addition to resolving the patients’ ALS symptoms—also treated the patients’ latent syphilis infections, as all five patients were negative for syphilis following treatment. Furthermore, the observation that syphilitic ALS has only been reported in Morocco may be explained by the striking fact that at the time, approximately 10% of the population in Morocco was positive for syphilis5. Therefore, these observations appeared to support the notion that syphilis can cause ALS, giving rise to the syphilitic ALS hypothesis.\n\nHere, however, I cast doubt on the notion that syphilitic ALS is a bona fide form of ALS and suggest that syphilis may in fact be a confounding factor. First, if syphilis can cause ALS, one would expect the prevalence of ALS to be higher in countries in which syphilis is highly prevalent. However, this has not been observed. During the period in which the five cases of syphilitic ALS were reported, the prevalence of syphilis was 300-fold higher in Morocco than in developed countries5, yet the prevalence of ALS was similar between Morocco and the developed countries.\n\nA second observation is that the ALS symptoms in these five patients either stabilized or resolved after the patients received increasing doses of penicillin G and hydrocortisone. From this perspective, it is important to note that both penicillin G and hydrocortisone have pharmacological profiles that extend beyond their pharmacotherapeutic applications for treating syphilis. For example, in addition to its antibacterial properties, penicillin G is also a potent antagonist of GABAergic (i.e., inhibitory) activity6 and can induce seizures in patients with renal insufficiency7. Interestingly, GABAergic overstimulation was postulated recently to play a role in the pathogenesis of ALS8. Furthermore, at sufficient dosages, penicillin G can affect several major bodily functions and/or systems, including the immune system, the cardiovascular system, metabolic function, renal function, liver function, the hematological system, and the urogenital system7.\n\nSimilarly, hydrocortisone is a known immunomodulatory and anti-inflammatory agent9, thus potentially affecting systems that are involved in the pathogenesis of ALS1–3. Moreover, hydrocortisone has reported efficacy in multiple sclerosis and respiratory diseases9, conditions that have clinical overlap with ALS. Furthermore, like penicillin, hydrocortisone can affect several bodily functions and/or systems, including the endocrine system, the immune system, inflammatory function, the respiratory system, the hematological system, and the gastrointestinal system9. Furthermore, hydrocortisone has been reported to affect the GABAergic system10, the GABAergic system has been reported to affect the release of cortisone, the natural form of hydrocortisone11. and glucocorticoids have been shown to be efficacious in preclinical models for ALS12.\n\nA third key observation is that syphilis represents a major confounding factor with respect to the interpretation of clinical outcome, particularly in countries in which syphilis is highly prevalent. The neurology center where the five cases of syphilitic ALS were identified in the 1980s reported that these five patients were identified from a total of 40 patients with ALS. This corresponds to a 12.5% prevalence of syphilis among ALS patients in this neurology center. Strikingly, this percentage is on par with the prevalence of syphilis in the general population of Morocco (i.e., 10%) in that same time period5. Given this extremely high prevalence, syphilis (particularly neurosyphilis) is likely a confounding factor in the interpretation of many clinical conditions, including neuromuscular diseases such as ALS.\n\nBased on these observations and potential confounding factors, interpreting the clinical outcome in these five patients with syphilitic ALS can lead to incorrect conclusions, unless this treatment is also tested in ALS patients who are negative for syphilis. Given that this has not been attempted, the potential confounding role of syphilis in syphilitic ALS allows for the possibility of alternative explanations. Therefore, I hypothesize that syphilis is actually a confounding factor in syphilitic ALS, and the apparent treatment of syphilitic ALS is actually due to the coincident treatment of both syphilis and ALS due to two separate therapeutic actions of the treatment regimen. I therefore propose that penicillin G and/or hydrocortisone may be effective at treating both syphilitic ALS and non-syphilitic ALS. If correct, this may have long-reaching implications for the treatment of ALS, a currently incurable disease.\n\nSupport for this hypothesis comes from a 2013 report of a patient with syphilitic ALS13. Given the relatively brief follow-up of this patient (less than one year), this case was not included in the current analysis. However, following a treatment regimen similar to treatment applied in the previous five cases, this patient was reportedly cured of ALS.\n\nOther explanations may also account for the observations discussed above. First, some of the five initial patients who were treated for syphilitic ALS may have been misdiagnosed. However, this is unlikely, given that the neurology center had extensive experience diagnosing both ALS and syphilis4. Furthermore, the symptoms associated with ALS1–3 generally do not overlap with the symptoms associated with syphilis5. Secondly, it is possible that the patients’ ALS symptoms were caused by an infection other than syphilis, and this other infection was treated by the penicillin and hydrocortisone. Moreover, the presence of a previously unidentified infection—and its treatment with penicillin G—may explain the observation that the 21-day treatment (see Table 1) provided long-term reversal of symptoms (i.e., up to 13 years). However, this is unlikely, as it would mean that all five patients had the same infection in addition to testing positive for syphilis. Most importantly, penicillin (a GABA receptor antagonist) may provide long-term treatment of ALS through an action other than its antibacterial activity. Precedence for this hypothesis comes from reports that other GABA receptor antagonists provide long-term effects by resetting the master circadian clock14. Moreover, hydrocortisone is also reported to affect circadian rhythm15.\n\nIn summary, given the published clinical observations, I propose that syphilis may actually be a confounding factor—not a causative agent—in the clinical entity currently called syphilitic ALS. The notion that syphilis may be a confounding factor in syphilitic ALS is highly relevant, as it suggests that treating patients with ALS—regardless of whether they have syphilitic ALS or non-syphilitic ALS—may be effective at slowing or even reversing the progression of ALS. Based on the rapidly progressive and highly devastating nature of ALS, further research is needed to test this novel hypothesis, possibly leading to the first effective treatment for ALS and perhaps other progressive neuromuscular diseases.",
"appendix": "Competing interests\n\n\n\nBT is the inventor on a patent application on the treatment of neuromuscular and neurologic diseases with therapies described in the manuscript, and founded Ry Pharma, a company that aims to develop such therapies.\n\n\nGrant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nReferences\n\nKiernan MC, Vucic S, Cheah BC, et al.: Amyotrophic lateral sclerosis. Lancet. 2011; 377(9769): 942–55. PubMed Abstract | Publisher Full Text\n\nSilani V, Messina S, Poletti B, et al.: The diagnosis of Amyotrophic lateral sclerosis in 2010. Arch Ital Biol. 2011; 149(1): 5–27. PubMed Abstract | Publisher Full Text\n\nTurner MR, Hardiman O, Benatar M, et al.: Controversies and priorities in amyotrophic lateral sclerosis. Lancet Neurol. 2013; 12(3): 310–22. PubMed Abstract | Publisher Full Text\n\nel Alaoui-Faris M, Medejel A, al Zemmouri K, et al.: [Amyotrophic lateral sclerosis syndrome of syphilitic origin. 5 cases.] Rev Neurol (Paris). 1990; 146(1): 41–4. PubMed Abstract\n\nSingh AE, Romanowski B: Syphilis: Review with Emphasis on Clinical, Epidemiologic, and Some Biologic Features. Clin Microbiol Rev. 1999; 12(2): 187–209. PubMed Abstract | Free Full Text\n\nRossokhin AV, Sharonova IN, Bukanova JV, et al.: Block of GABAA receptor ion channel by penicillin: electrophysiological and modeling insights toward the mechanism. Mol Cell Neurosci. 2014; 63: 72–82. PubMed Abstract | Publisher Full Text\n\nPenicillin G summary of product characteristics. Accessed July 16, 2016. Reference Source\n\nTuk B: Inhibitory system overstimulation plays a role in the pathogenesis of neuromuscular and neurological diseases: a novel hypothesis [version 1; referees: awaiting peer review]. F1000Res. 2016; 5: 1435. Publisher Full Text\n\nHydrocortisone Summary of product characteristics. Accessed July 16, 2016. Reference Source\n\nOng J, Kerr DI, Johnston GA: Cortisol: a potent biphasic modulator at GABAA-receptor complexes in the guinea pig isolated ileum. Neurosci Lett. 1987; 82(1): 101–6. PubMed Abstract | Publisher Full Text\n\nCullinan WE: GABAA receptor subunit expression within hypophysiotropic CRH neurons: a dual hybridization histochemical study. J Comp Neurol. 2000; 419(3): 344–51. PubMed Abstract | Publisher Full Text\n\nEvans MC, Gaillard PJ, de Boer M, et al.: CNS-targeted glucocorticoid reduces pathology in mouse model of amyotrophic lateral sclerosis. Acta Neuropathol Commun. 2014; 2: 66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChraa M, Mebrouk Y, McCaughey C, et al.: Amyotrophic lateral sclerosis mimic syndrome due to neurosyphilis. Amyotroph Lateral Scler Frontotemporal Degener. 2013; 14(3): 234. PubMed Abstract | Publisher Full Text\n\nRuby NF, Hwang CE, Wessells C, et al.: Hippocampal-dependent learning requires a functional circadian system. Proc Natl Acad Sci U S A. 2008; 105(40): 15593–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChan S, Debono M: Replication of cortisol circadian rhythm: new advances in hydrocortisone replacement therapy. Ther Adv Endocrinol Metab. 2010; 1(3): 129–138. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "15388",
"date": "22 Aug 2016",
"name": "Pieter Gaillard",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDr Bert Tuk has provided the readers with a fascinating Opinion Article, in which he sets out his hypothesis that syphilitic ALS is actually not caused by syphilis. And as a result, the pharmacotherapy that was used to treat and cure the confounding syphilitic origin of the ALS symptoms may offer the first effective treatment option for patients suffering from the fatal disease ALS, be it with or without the presence of syphilis.\n\nThis Opinion Article builds on the earlier presented Opinion Article in this journal by Dr. Tuk, entitled: Inhibitory system overstimulation plays a role in the pathogenesis of neuromuscular and neurological diseases: a novel hypothesis1. In this paper he discloses his assessment of a fundamentally new view of how the symptoms of ALS could be caused, essentially by an overstimulation of the GABA neurotransmitter system, and his current paper basically provided case study proof of concept for this hypothesis with a specific treatment regimen of existing drugs (penicillin G and hydrocortisone) that may indeed act to normalise this overstimulated GABA neurotransmitter system.\n\nThe methods and analysis of the existing literature and data are sound, as is the discussion leading to the new hypothesis, although a more direct link and explanation to the suggested mechanism (modification of the overstimulation of the GABA neurotransmitter system) would have been helpful to the reader.\n\nAnd a few more words of caution could have been added to the discussion section, since this work will likely result in a “call to action” by patients and their treating physicians. It is therefore important to advise to keep tight involvement of expert (clinical) pharmacologists, neurologists, and neuroscientists as it is yet unknown what the impact of any modification to this specific treatment regimen would do to an individual patient. Today we don’t know how strict the applied pharmacotherapeutic regimen is to render the expected effect. Can the regimen be modified, or improved? Will modification in the regimen result in any effects on tolerance development, or perhaps sensitization, with risk of seizure activity? How will the two drugs (penicillin G and hydrocortisol) interact when the regimen is modified? Can other compounds be used with a similar pharmacologic profile? Definitely one cannot simply replace to other antibiotics, like minocycline, which was proven effective as alternative in treating neurosyphilis2, yet in ALS patients proven to be detrimental3, perhaps - according the hypothesis of Dr. Tuk -because minocycline has GABA agonistic properties4. Many key questions that deserve careful consideration, monitoring and further studies. Other domains that require further attention is the role of the bulbar onset and treatment effect, since this patient seemed less responsive. And since the two drugs effect two interconnected major systems in the body simultaneously (penicillin G on the inhibitory neurotransmitter system and hydrocortisol on the endocrinology system, both centrally and systemically), there is a world of knowledge to be gained in ALS and also other related diseases how this interaction will play out, leading to hopefully to the long awaited improvements for these devastating brain diseases.",
"responses": []
},
{
"id": "16950",
"date": "21 Oct 2016",
"name": "Alan Gill",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction\n\nThe antecedent hypothesis for this work might be summarized in this way: excitatory overstimulation is a homeostatic response to inhibitory overstimulation. What is actually needed is to reduce inhibitory activity (the primary problem) so compensatory excitatory overstimulation is no longer required.\n\nGiven the widely accepted working model that reduced inhibitory and excessive excitatory neuronal activity contributes to ALS through neuronal excitotoxicity, the hypothesis asks that we look at ALS with new eyes. In this devastating disease for which we have no satisfactory treatment, looking with new eyes is certainly warranted.\n\nThe current work provides evidence that co-treatment with PENG and hydrocortisone beneficially slows ALS progression, even in ALS patients that did not have syphilis. The current work’s hypothesis might be summarized in this way: PENG helps ALS symptoms because it is a GABA receptor antagonist. Hydrocortisone helps with inflammation also present in ALS.\n\nApart from the limitations of our understanding ALS pathology, the current work opens an array of complexities, including regulatory interneuron networks, inhibitory receptor subtypes, antagonist drug properties and side-effects, neuronal ion channel composition, neuropeptide modulation, central pattern generator function, and neuroinflammation. For example, why tiptoe into GABA antagonism with PENG when more refined and specific inhibitors are available. Or, patient-to-patient variability in blood-brain barrier leakiness could make it difficult to control CNS PENG levels and the probability of seizure induction. Or, should somatostatin-expressing interneuron networks be selected for in our treatment approach? Should studies include cohorts that do not receive hydrocortisone, or only hydrocortisone? Combination drug studies, if anything is to be learned from them, require expensive designs that test each drug separately and the identification of a basis for doses used in the combination, or combinations.\n\nThe complexity is formidable. But in the end, what is really important is whether PENG plus hydrocortisone actually slows ALS progression, i.e. “does it work?”. Even if we don’t understand why we have an obligation to try what works. The true purpose of understanding the complexities above is to try to design a reliable therapeutic regimen for clinical application. We want the therapy to work predictably and we don’t want it to do harm.\n\nThe comments and references that I have compiled below are intended to stimulate the process of making the potential therapy safe and predictable.\n\nComments and References:\n\nThe outcome of homeostatic regulation depends on the complement of ion channels expressed in cells: in some cases, loss of specific ion channels can be compensated; in others, the homeostatic mechanism itself causes pathological function1.\n\nThere may be a need to target restricted GABAergic subpopulations and cell types characterized by distinct laminar location, morphology, axonal projection, and electrophysiological properties2.\n\nIncreasing GABA May Be Helpful\n\nThe GABA analog Pregabalin increases extracellular GABA by inducing its synthesis and transport. Cacna2d2, the gene encoding the Alpha2delta2 subunit of voltage-gated calcium channels (VGCCs), is a developmental switch that limits axon growth and regeneration. In vivo, Alpha2delta2 pharmacological blockade through Pregabalin administration enhanced axon regeneration in adult mice after spinal cord injury. Since Pregabalin is already an established treatment for a wide range of neurological disorders, targeting Alpha2delta2 may be a novel treatment strategy to promote structural plasticity and regeneration following CNS trauma3.\n\nElectrophysiological and histological studies support the pathophysiological concept of an impaired inhibitory, namely GABAergic, control of the motoneurons in the cerebral cortex of ALS patients. However, in the prefrontal and temporal cortex of ALS patients the GABA synthesizing enzyme glutamic acid decarboxylase (GAD) was significantly upregulated4.\n\nThere are reduced levels of GABA in the motor cortex of patients with ALS. Patients with ALS had significantly lower levels of GABA in the motor cortex than did healthy controls. Riluzole-naive patients with ALS had higher levels of glutamate-glutamine than did riluzole-treated patients with ALS or healthy controls5,6.\n\nIncreasing Glycinergic Function May Be Helpful\n\nThe selective loss of inhibitory glycinergic regulation of motoneuron function or glycinergic interneuron degeneration contributes to motoneuron degeneration in amyotrophic lateral sclerosis7.\n\nAnti-glutamate drugs don’t have major disease-modifying effects clinically in ALS. Nevertheless, hyperexcitability of upper and lower motor neurons is a feature of human ALS and transgenic (tg) mouse models of ALS. Motor neuron excitability is modulated by presynaptic glycinergic and GABAergic innervation and postsynaptic glycine and GABA(A) receptors that mediate synaptic inhibition. Inhibitory glycinergic innervation of spinal motor neurons becomes deficient before motor neuron degeneration is evident in G93A-hSOD1 mice. GABAergic innervation of ALS mouse motor neurons and GABA(A) receptor function appear normal. Abnormal synaptic inhibition resulting from dysfunction of interneurons and motor neuron Glycine receptors may participate in ALS pathogenesis8.\n\nMuscle paralysis during REM sleep requires activation of both GABA and glycine receptors together. Glycine alone doesn’t paralyze the muscles; GABA is also required. To prevent muscle paralysis during REM sleep both types of receptors have to be off. GABA antagonists like bicuculline, metrazol, and flumazenil would remove the GABA effect and reverse paralysis, stimulating muscle movement. In general these drugs produce stimulant and convulsant effects, and are mainly used for counteracting overdoses of sedative drugs. There is no clear way to direct their activity in a specific direction to solve ALS paralysis. GABA interneurons are part of most neuronal networks in the brain and spinal cord. They act to modulate activity through the networks and they do so by monitoring activity level and responding to the level in a way that keeps it within normal limits. They participate in rhythm and pattern generating networks that coordinate many kinds of brain and spinal cord activity, including breathing and movement.\n\nGABA loss seems to be less influential than glycine loss. In mice lacking physiological levels of GABA, there are distinct regional changes in motor neuron number and muscle innervation, which appear to be linked to their physiological function and to their rostral-caudal position within the developing spinal cord. For more caudal (lumbar) regions of the spinal cord, the effect of GABA is less influential on motor neuron development compared to that of glycine9.\n\nGABA(A) and glycine receptors are the major inhibitory neurotransmitter receptors and contribute to many synaptic functions. Gephyrin promotes recruitment of GABA(A) receptors instead of glycine receptors. This mechanism could affect the balance of GABA(A) vs. glycine in ALS10.\n\nReduced short interval intracortical inhibition that leads to development of cortical hyperexcitability in ALS represents degeneration of inhibitory cortical circuits, combined with excessive excitation of high threshold excitatory pathways. Neuroprotective strategies aimed at preserving the integrity of intracortical inhibitory circuits, in addition to antagonizing excitatory cortical circuits, may provide novel therapeutic targets in ALS11.\n\n“A wide range of evidence from clinical, histological, genetic, neurophysiological, neuroimaging and neuropsychological studies, suggests that a loss of central nervous system inhibitory neuronal influence is a contributing factor in ALS pathogenesis. This loss of inhibitory function points intuitively to an 'interneuronopathy', with natural differences in cortical and spinal inhibitory networks reflected in the hitherto unexplained variable compartmentalization of pathology within upper and lower motor neuron populations. An excitotoxic final common pathway might then result from unopposed glutamatergic activity. If correct, therapies aimed specifically at supporting interneuronal function may provide a novel therapeutic strategy.”12.\n\nAlong with GABA and glycine, the coexpressed interneuron neuromodulatory peptides, like somatostatin, VIP, CCK, and parvalbumin have major influences on the network functions and channel expression. Other neuropeptides also contribute to the complexity. α-MSH preserves GAD67 expression and prevents loss of the somatostatin-expressing subtype of GABAergic GAD67+ inhibitory interneurons13.\n\nGABAergic Interneurons Contribute to Central Pattern Generators\n\nWell controlled pharmacologic manipulation of GABAergic networks may be difficult.\n\nCentral pattern generators are neuronal circuits that when activated can produce rhythmic motor patterns such as walking, breathing, flying, and swimming in the absence of sensory or descending inputs that carry specific timing information14.\n\nA class of GABAergic sensory neurons contacting the CSF in the vertebrate spinal cord are referred to as CSF-cNs. These cells modulate components of locomotor central pattern generators. CSF-cNs form active GABAergic synapses onto V0-v glutamatergic interneurons, an essential component of locomotor Central Pattern Generators. Spinal GABAergic sensory neurons can tune the excitability of locomotor Central Pattern Generators in a state-dependent manner by projecting onto essential components of the excitatory premotor pool15.\n\nMotor neurons themselves can control, by retrograde transmission, upstream glutamatergic V2a interneurons that drive locomotion16.\n\nThe lumbar central pattern generator for locomotion selectively modulates the intracellular activity of spinal respiratory neurons engaged in expiration. Medullary respiratory network activity of both thoracolumbar expiratory motor neurons and interneurons is rhythmically modulated with hindlimb ipsilateral flexor locomotor activity. Locomotion and respiration must to be tightly coordinated to reduce muscular interferences and facilitate breathing rate acceleration during exercise17.\n\nConclusion\n\nThe question remains. Can we identify a safer, more predictable, therapeutic regimen than PENG/Hydrocortisone by trying to understand which inhibitory interneuron networks require interventions and which do not? And which specific antagonist (or agonist) drugs would optimize our test of the hypothesis? Can specific peptide neuromodulators contribute to improving the regimen? These considerations do not preclude the importance of carefully testing the PENG/Hydrocortisone regimen itself. The complexity of this task alone is significant. Refinements of the regimen only add complexity, but may make the difference between safe, practical use in ALS and early rejection of the approach.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1904
|
https://f1000research.com/articles/4-157/v1
|
18 Jun 15
|
{
"type": "Review",
"title": "Drosophila's contribution to stem cell research",
"authors": [
"Gyanesh Singh"
],
"abstract": "The discovery of Drosophila stem cells with striking similarities to mammalian stem cells has brought new hope for stem cell research. A recent development in Drosophila stem cell research is bringing wider opportunities for contemporary stem cell biologists. In this regard, Drosophila germ cells are becoming a popular model of stem cell research. In several cases, genes that controlled Drosophila stem cells were later discovered to have functional homologs in mammalian stem cells. Like mammals, Drosophila germline stem cells (GSCs) are controlled by both intrinsic as well as external signals. Inside the Drosophila testes, germline and somatic stem cells form a cluster of cells (the hub). Hub cells depend on JAK-STAT signaling, and, in absence of this signal, they do not self-renew. In Drosophila, significant changes occur within the stem cell niche that contributes to a decline in stem cell number over time. In case of aging Drosophila, somatic niche cells show reduced DE-cadherin and unpaired (Upd) proteins. Unpaired proteins are known to directly decrease stem cell number within the niches, and, overexpression of upd within niche cells restored GSCs in older males also . Stem cells in the midgut of Drosophila are also very promising. Reduced Notch signaling was found to increase the number of midgut progenitor cells. On the other hand, activation of the Notch pathway decreased proliferation of these cells. Further research in this area should lead to the discovery of additional factors that regulate stem and progenitor cells in Drosophila.",
"keywords": [
"Stem cell",
"Drosophila",
"Invertebrate stem cell"
],
"content": "Introduction\n\nThe fundamental property of stem cells that they can not only differentiate into various types of cells, but can also renew the stem cell population, is the basis of progressive regenerative medicine1. It is normal for some tissues like blood, skin, gut and germ cells to be regularly maintained by stem cell precursors. Stem cell niches control important properties of stem cells including self-renewing potential2. Currently, Drosophila germ cells are established as a crucial model of stem cells.\n\n\nAnalysis of the recent literature\n\nDrosophila ovary contains both germline and somatic stem cells that reside within the anterior region of each ovariole3. In an interesting experiment, where individual germaria, free of developing eggs and sheath tissue, were transplanted into the abdominal cavity of a host Drosophila, they not only regenerated ovariole-like structures but also maintained oogenesis4. Drosophila ovariole usually contains two somatic stem cells (called cystocysts) near the wall of the germarium. Interestingly, somatic stem cells, in this case, not only divide independently of surrounding cells, but also continue to divide in the absence of germline cells5. As asymmetric stem cell division is an important property that enables stem cells to self-renew and differentiate, the balance between symmetry and asymmetry is a tool that enables stem cells to maintain required numbers of progeny cells. An enormous amount of research effort has been directed towards understanding the basis of this asymmetry. Ablation of presumptive germline stem cells (GSCs) near the apical tip blocked the production of new germline cysts, however, previously initiated cysts were able to complete development in this case6. This indicated that development of cysts does not require continued cyst production. More importantly, ablation of a distinct group of somatic cells around GSCs leads to higher egg production7. It has been reported that stem cells adjust their proliferation rate in response to nutrition without changing the number of active stem cells e.g. a protein-rich diet increases the rate of egg production, in this case8.\n\nGermline and somatic stem cells attach to form a cluster of cells (the hub) in the Drosophila testes. The hub expresses a ligand that activates the JAK-STAT signaling cascade9. Without this signal, GSCs do not self-renew, but can differentiate. Drosophila bag of marbles (bam) gene is required for the differentiation of daughter cells (cystoblasts) from mother stem cells10. Instead of differentiation, bam mutant germ cells proliferated like stem cells. Heat-induced bam expression caused elimination of germinal stem cells while somatic stem cell numbers were not changed11. Interestingly, ectopic bam expression had no such consequences on male germline cells indicating bam’s potential to regulate oogenesis and spermatogenesis in different ways11,12. Somatic cyst cells and hub cells express two bone morphogenetic protein (BMP) molecules: Gbb (Glass bottom boat) and Dpp (Decapentaplegic). The Dpp/BMP signal was found to be essential for GSC maintenance13. In absence of BMP signaling, bam is upregulated that can cause GSCs to be lost. Mutations in Dpp or its receptor (saxophone) increases stem cell loss and inhibits stem cell division. On the other hand, overexpression of Dpp blocks GSC differentiation13. Interestingly, BMP signaling reduces bam expression in ovarian GSCs. Phosphorylated Mad (pMad) is an indirect indicator of BMP signaling as it upregulates in response to repressed bam expression14. Somatic inner germarium sheath cells failed to divide after removing GSC niches. Hedgehog (Hh) family signaling mediators are known for their important role during Drosophila development15. Hedgehog genes were also reported to be crucial for the proliferation of ovarian somatic cells in Drosophila. Drosophila neuroblasts regulate stem cell growth by separating the growth inhibitor Brat and the transcription factor Prospero into different daughter cells15. Interestingly, mutant Brat or Prospero caused both daughter cells to grow resulting into tumorigenesis16. High levels of Pumilio and Nanos proteins have also been observed in Drosophila GSCs17. Lack of zygotic activity of Nanos or Pumilio was found to have a dramatic effect on germline development in female flies. Pumilio mutant Drosophila not only failed to maintain stem cells but germline cells also17. Loqs protein was also found to be necessary for embryo survival and GSC sustenance in Drosophila. Decrease in stem cell functions could lead to the aging-related decline in tissue maintenance18,19. Somatic niche cells in testes from aging males show reduced DE-cadherin and unpaired (Upd) proteins20. Inside the Drosophila testes, Upd production in hub cells controls stem cell number within the niches, and overexpression of upd within niche cells can rescue GSCs even in case of aged males.\n\nThe identification of stem cell lineages in the midgut of Drosophila is a recent discovery21. A genome-wide transgenic RNAi screen identified 405 genes that regulate intestinal stem cell (ISC) maintenance and differentiation in Drosophila intestine22. By integrating these genes into functional networks, it was concluded that factors related to basic stem cell processes are commonly needed in all stem cells, and stem-cell-specific, niche-related signals are required only in the unique stem cell types. Analysis of genetic mosaics revealed that differentiated cells in the midgut epithelium come from a common lineage in Drosophila23. Notch signaling controls key events during development. Consistent with its role of regulation of various adult stem cells, diminished notch signaling has been reported to cause increase in the number of precursor cells in the midgut of Drosophila24.\n\n\nConclusions\n\nDrosophila germline and midgut stem cells are being established as important models of stem cell research. Self-renewal of GSCs requires both intracellular as well as extracellular signals. Several factors including BMP signals were found to be indispensable for sustaining GSCs in Drosophila. Asymmetric division of GSCs to produce and maintain a daughter GSC is regulated by gene expression in adjacent somatic cells also. In Drosophila, significant changes occur within the stem cell niche that contributes to a decline in stem cell number over time. For more than a century, Drosophila’s contribution to genetics and developmental biology has been enormous. With its increasing contribution to stem cell research, Drosophila consistently proves to be an invaluable model organism.",
"appendix": "Author contributions\n\n\n\nGS conceived the study and prepared the first draft of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe author is thankful to Rohit Sood and Manpreet Kaur for their valuable suggestions.\n\n\nReferences\n\nSomorjai IM, Lohmann JU, Holstein TW, et al.: Stem cells: a view from the roots. Biotechnol J. 2012; 7(6): 704–22. PubMed Abstract | Publisher Full Text\n\nPapagiannouli F, Lohmann I: Shaping the niche: lessons from the Drosophila testis and other model systems. Biotechnol J. 2012; 7(6): 723–36. PubMed Abstract | Publisher Full Text\n\nEliazer S, Buszczak M: Finding a niche: studies from the Drosophila ovary. Stem Cell Res Ther. 2011; 2(6): 45. PubMed Abstract | Publisher Full Text\n\nZhao R, Xuan Y, Li X, et al.: Age-related changes of germline stem cell activity, niche signaling activity and egg production in Drosophila. Aging Cell. 2008; 7(3): 344–354. PubMed Abstract | Publisher Full Text\n\nXie T, Spradling AC: A niche maintaining germ line stem cells in the Drosophila Ovary. Science. 2000; 290(5490): 328–330. PubMed Abstract | Publisher Full Text\n\nLin H, Spradling AC: Germline stem cell division and egg chamber development in transplanted Drosophila germaria. Dev Biol. 1993; 159(1): 140–152. PubMed Abstract | Publisher Full Text\n\nChen D, McKearin DM: A discrete transcriptional silencer in the bam gene determines asymmetric division of the Drosophila germline stem cell. Development. 2003; 130(6): 1159–1170. PubMed Abstract | Publisher Full Text\n\nDrummond-Barbosa D, Spradling AC: Stem cells and their progeny respond to nutritional changes during Drosophila oogenesis. Dev Biol. 2001; 231(1): 265–278. PubMed Abstract | Publisher Full Text\n\nSinden D, Badgett M, Fry J, et al.: Jak-STAT regulation of cyst stem cell development in the Drosophila testis. Dev Biol. 2012; 372(1): 5–16. PubMed Abstract | Publisher Full Text\n\nPerinthottathil S, Kim C: Bam and Bgcn in Drosophila germline stem cell differentiation. Vitam Horm. 2011; 87: 399–416. PubMed Abstract | Publisher Full Text\n\nOhlstein B, McKearin D: Ectopic expression of the Drosophila Bam protein eliminates oogenic germline stem cells. Development. 1997; 124(18): 3651–3662. PubMed Abstract\n\nAberle H, Haghighi AP, Fetter RD, et al.: Wishful thinking encodes a BMP type II receptor that regulates synaptic growth in Drosophila. Neuron. 2002; 33(4): 545–558. PubMed Abstract | Publisher Full Text\n\nGibson MC, Perrimon N: Extrusion and death of DPP/BMP-compromised epithelial cells in the developing Drosophila wing. Science. 2005; 307(5716): 1785–1789. PubMed Abstract | Publisher Full Text\n\nTanimoto H, Itoh S, ten Dijke P, et al.: Hedgehog creates a gradient of DPP activity in Drosophila wing imaginal discs. Mol Cell. 2000; 5(1): 59–71. PubMed Abstract | Publisher Full Text\n\nMichel M, Kupinski AP, Raabe I, et al.: Hh signalling is essential for somatic stem cell maintenance in the Drosophila testis niche. Development. 2012; 139(15): 2663–9. PubMed Abstract | Publisher Full Text\n\nBetschinger J, Mechtler K, Knoblich JA: Asymmetric segregation of the tumor suppressor brat regulates self-renewal in Drosophila neural stem cells. Cell. 2006; 124(6): 1241–1253. PubMed Abstract | Publisher Full Text\n\nJaruzelska J, Kotecki M, Kusz K, et al.: Conservation of a Pumilio-Nanos complex from Drosophila germ plasm to human germ cells. Dev Genes Evol. 2003; 213(3): 120–126. PubMed Abstract | Publisher Full Text\n\nSousa-Victor P, García-Prat L, Serrano AL, et al.: Muscle stem cell aging: regulation and rejuvenation. Trends Endocrinol Metab. 2015; 26(6): 287–296. PubMed Abstract | Publisher Full Text\n\nSuksuphew S, Noisa P: Neural stem cells could serve as a therapeutic material for age-related neurodegenerative diseases. World J Stem Cells. 2015; 7(2): 502–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang H, Singh SR, Zheng Z, et al.: Rap-GEF signaling controls stem cell anchoring to their niche through regulating DE-cadherin-mediated cell adhesion in the Drosophila testis. Dev Cell. 2006; 10(1): 117–126. PubMed Abstract | Publisher Full Text\n\nde Navascués J, Perdigoto CN, Bian Y, et al.: Drosophila midgut homeostasis involves neutral competition between symmetrically dividing intestinal stem cells. EMBO J. 2012; 31(11): 2473–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZeng X, Han L, Singh SR, et al.: Genome-wide RNAi screen identifies networks involved in intestinal stem cell regulation in Drosophila. Cell Rep. 2015; 10(7): 1226–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJiang H, Edgar BA: Intestinal stem cell function in Drosophila and mice. Curr Opin Genet Dev. 2012; 22(4): 354–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPerdigoto CN, Bardin AJ: Sending the right signal: Notch and stem cells. Biochim Biophys Acta. 2013; 1830(2): 2307–22. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "9120",
"date": "29 Jun 2015",
"name": "Takashi Adachi-Yamada",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI would like to accept this short review after the author makes improvements described below. For better contrast to the following sentence, the word “female” should be inserted between the “elimination of ” and “germinal stem cells” in the sentence “Heat-induced bam expression caused elimination of germinal stem cells while somatic stem cell numbers were not changed” in the third section. The author should reconfirm that the references 12 and 13 are appropriately cited. The author stated that Phosphorylated Mad (pMad) is an indirect indicator of BMP signaling. However, unlike the case of various reporter genes, I think that pMad is a “direct” indicator because it is directly phosphorylated by BMP type I receptors. At the position of citation of reference #21, the author should also cite two original papers that first described Drosophila intestinal stem cells, i.e. Ohlstein & Spradling (2006) and Micchelli and Perrimon (2006). It would be nice if the author concretely indicate some representative factors that are commonly used in all stem cells at the position of reference #22. I think that most of readers in this field have great interest in this recent discovery.",
"responses": []
},
{
"id": "9121",
"date": "30 Jun 2015",
"name": "Surajit Sarkar",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the manuscript entitled “Drosophila's contribution to stem cell research” by Gyanesh Singh; the author provides an overview of the stem cell research in Drosophila. The manuscript provide a brief survey of the recent findings and discuss about various signalling pathways operating in germline stem cell niche. Though it is a good reading and loaded with quality scientific information, some minor corrections may be incorporated as follow:The writing may be improved at some places such as, in second line of abstract “A recent development” should be “Recent developments”. Conclusions of the manuscript may be improved. It should provide a clear and concluding message to the readers.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/4-157
|
https://f1000research.com/articles/5-1141/v1
|
03 Jun 16
|
{
"type": "Software Tool Article",
"title": "Biomedical Mutation Analysis (BMA): A software tool for analyzing mutations associated with antiviral resistance",
"authors": [
"Karina Salvatierra",
"Hector Florez",
"Karina Salvatierra"
],
"abstract": "Introduction: Hepatitis C virus (HCV) is considered a major public health problem, with 200 million people infected worldwide. The treatment for HCV chronic infection with pegylated interferon alpha plus ribavirin inhibitors is unspecific; consequently, the treatment is effective in only 50% of patients infected. This has prompted the development of direct-acting antivirals (DAA) that target virus proteins. These DAA have demonstrated a potent effect in vitro and in vivo; however, virus mutations associated with the development of resistance have been described. Objective: To design and develop an online information system for detecting mutations in amino acids known to be implicated in resistance to DAA. Materials and methods:\n\nWe have used computer applications, technological tools, standard languages, infrastructure systems and algorithms, to analyze positions associated with resistance to DAA for the NS3, NS5A, and NS5B genes of HCV. Results: We have designed and developed an online information system named Biomedical Mutation Analysis (BMA), which allows users to calculate changes in nucleotide and amino acid sequences for each selected sequence from conventional Sanger and cloning sequencing using a graphical interface. Conclusion: BMA quickly, easily and effectively analyzes mutations, including complete documentation and examples. Furthermore, the development of different visualization techniques allows proper interpretation and understanding of the results. The data obtained using BMA will be useful for the assessment and surveillance of HCV resistance to new antivirals, and for the treatment regimens by selecting those DAA to which the virus is not resistant, avoiding unnecessary treatment failures. The software is available at: http://bma.itiud.org.",
"keywords": [
"Hepatitis C virus",
"mutations",
"resistance",
"information system"
],
"content": "Introduction\n\nChronic hepatitis C infection is caused by the hepatitis C virus (HCV) and affects an estimated 200 million people worldwide1,2. Transmission occurs by percutaneous exposure through blood products. The major risk factors for HCV infection are parenteral exposure, and needle sharing among intravenous drug users. In addition, hemodialysis patients are at risk of contracting an HCV infection3–5.\n\nHistorically, HCV drug therapy has depended on interferon-α and ribavirin and the effectiveness of this combination therapy are primarily determined by the HCV genotype6. The advent of direct-acting antiviral agents (DAA) has paved the way for a new era for the treatment of HCV infection. The most important contribution in their development primarily target protease NS3, protein NS5A or NS5B RNA-dependent RNA polymerase7. However, because the emergence of resistant viral variants, DAA is one of the factors to be taken into account in the treatment8. Antiviral capacity may be limited by the ability of the virus to develop resistance to new antivirals9. Resistance mutations to DAA have been observed both in vitro and in vivo10–12. In addition, people infected with HCV, who are left untreated, can develop natural viral variants harboring resistance mutations. Current data indicates pre-existing mutations to NS3 protease inhibitors, NS5A inhibitors, and non-nucleoside inhibitors of NS5B polymerase in 7.7%, 16.2% and 22.5%, of infected patients13–16. Probably these viral variants contribute to the selection of resistance to DAA during the initial weeks of monotherapy17–20.\n\nUsing DAA implies the possibility of selection of resistant variants. Antiviral resistance results from amino acid substitutions that produce conformational changes that interfere with drug-target interaction. These mutations typically involve a biological cost, and viruses carrying these mutations are found in smaller numbers than wild-type viruses; however, they can be positively selected during therapy21.\n\nGenetic variability affects the response to old and new therapies. It is therefore important to determine mutations of resistance to antiviral drugs.\n\nThere is an increasing need to develop bioinformatic tools to analyze the rapidly growing amount of nucleotide and amino acid sequence data in different organisms such as viruses. An important task in bioinformatics is the provisioning of data and tools in a simple manner for users to locate and use. Sequencing generates large amounts of data that need to be analyzed. Advances in information technology have stimulated the development of new computer applications and algorithms for data analysis, and computer visualization tools for the representation of variation patterns. The analysis of mutations is important to understand antiviral resistance and to understand the functions of different proteins. The aim of this study was to develop an online information system named Biomedical Mutation Analysis (BMA), which allows users to calculate changes in nucleotide and amino acid sequences for each selected sequence through a graphical interface.\n\n\nMaterials and methods\n\nTo create the online information system, we used different standard tools, languages, and infrastructure systems. BMA was designed using the Unified Modeling Language (UML)23, which allows describing the system following the Object Oriented Paradigm. Regarding the development of BMA, we used PHP language version 5.3.29 (https://secure.php.net/), which is supported by Apache software version 2.4.7 (http://www.apache.org/) as the application server. For the front end of BMA, we used Bootstrap version 3.3.6 (http://getbootstrap.com/), which is the most popular HTML, CSS, and JavaScript framework for developing responsive web projects. BMA also has some features based on JavaScript language supported by JQuery version 1.12.3 (https://jquery.com/), which is a JavaScript library that facilitates some specific JavaScript functionalities. BMA provides three different outputs, where two of them use additional support. The former result is a report generated as a pdf file, which is built using ezpdf version 0.0.9 (https://github.com/rebuy-de/ezpdf), which is a library that supports the creation of pdf files. The latter result is a force-directed graph, which is created using D3 (Data-Driven Documents) version 3.5.16 (https://d3js.org/), which is an online JavaScript library that helps to deploy data using fancy visualizations.\n\nBMA stores all information related to the mutation analyses in one database supported by MySql version 5.7.12 (https://www.mysql.com/), which is a relational database management system. The database includes the entities and relationships required for handling all information related to the proposed mutation analyses. The database is manipulated through project phpMyAdmin version 4.3.11 (https://www.phpmyadmin.net/), which is software written in PHP intended to handle the administration of data stored in MySql databases.\n\nThe database was design using the tool MySql Workbench version 6.3 (https://www.mysql.com/products/workbench/), while the online system was developed using the tool Eclipse PHP version 3.7.0 (https://eclipse.org/pdt/). BMA is hosted in a Linux Server debian distribution version 8.4, which includes Apache, MySql, and phpMyAdmin for the right operation of BMA.\n\nAll software, frameworks, and libraries used in the design and development of BMA have a GNU General Public License (GNU GPL) (http://www.gnu.org/licenses/licenses.en.html), which implies that BMA was completely created using free software.\n\nWe used the nucleotide sequence of genes NS3, NS5A and NS5B of Con1 isolated HCV genotype 1b (accession number: AJ238799), extracted from GenBank (www.ncbi.nlm.nih.gov/genbank/) as a reference sequence.\n\nA compilation of resistance mutations previously described in vivo and in vitro in the literature for the genes NS3, NS5A and NS5B of the HCV were used for computing the number and type of amino acid variants at the corresponding positions associated with resistance to DAA24,25.\n\n\nResults\n\nThe BMA’s core is the analysis algorithm that is able to evaluate multiple patients, where each one can include multiple sequences. In addition, the algorithm can analyze desired positions that the analyst can define. The execution of the algorithm is just one part of the complete analysis process. The analysis process includes the following steps:\n\n1. The analyst accesses BMA via the web site and selects the option “HCV” from the “Mutation Analysis” menu. BMA presents the list of genes available for HCV, which includes the name, description, and reference sequence (by clicking on the corresponding icon). Figure 1 presents the list of available genes.\n\n2. The analyst can use the search icon placed in each gene of the HCV (e.g., NS3, NS5A, NS5B) to proceed to the following step, which corresponds to the selection of the positions to be analyzed. Thus, possible positions are sorted in a list, which includes the number of the position, mutation, antiviral, inhibitor, and references that can be in vitro or in vivo. It is important to mention that BMA is flexible allowing the inclusion of further positions, mutations, and antivirals established in new or future research. Regarding references, each position presents the list of academic papers that support scientifically the inclusion of the position in the mutation analysis. Furthermore, for each position, there is an icon that lists the reference details with a link that redirects to one academic search service with the information of the selected reference. Figure 2 presents the list of some positions for the gene NS3.\n\n3. After selecting the positions to perform the analysis, the analyst is asked to provide the patient sequences as plain text files. BMA offers an example dataset for testing the analysis. BMA can automatically read and analyze multiple data files sequentially. These data files may contain a varying number of sequences that represent one patient. BMA can recognize plain text files, but they have to follow a specific format (see Figure 3). Files must include the symbol '>' and the sequence name in the first line of the file. The sequence data starts on the second line. Nucleotide data must be written in one line. The sequence must include the symbols: A, C, G, T. Sequences can also include the symbol '-' for specifying missing data. In sequences, blank spaces, tabs, break lines and other symbols are not accepted.\n\n4. Once patient files are selected, the analysis algorithm is executed. The algorithm presents the results in three different ways:\n\na) Online textual visualization of necessary nucleotide changes that produce an amino acid change, which generates resistance (Figure 4).\n\nb) An automatically generated report, which is sent to the analyst’s e-mail address. This report contains a summary of the calculated mutations for each sequence and the full detailed report of the executed analysis (Figure 5).\n\nc) A “force-directed” graph that identifies mutations of each patient sequence through node grouping, which corresponds to each analyzed sequence (Figure 6).\n\nFor reliable calculations the sequences must contain a substantial part of the genes NS3, NS5A or NS5B.\n\nThe analysis algorithm is based on multiple iterations. It collects all patients’ plain text files and iterates in order to analyze all of them independently. For each plain text file, the algorithm collects all sequences. Later on, for each sequence, the algorithm performs a new iteration using the selected positions. Then, for each position, it compares the nucleotide and amino acid of the iterated patient sequence with the reference sequence in the iterated position. At this stage, the information about changes is collected with the corresponding patient, sequence, and position. By finishing the execution of the algorithm, BMA uses the collected results to provide the three aforementioned visualizations.\n\nIt is important to mention that BMA cannot align sequences. There are some programs that can do this. For example, CLUSTAL W26 allows multiple alignments. In addition, DNA sequences cannot be edited or manipulated by BMA. No clinical decision should be based only on the result of BMA.\n\n\nConclusions\n\nSoftware for the detection of mutations associated with resistance to new DAAs is an important tool, because it guarantees accurate and reliable results. Moreover, BMA is freely available, which is different from others such as Bioedit, VectorNTI or MEGA, because it not only allows researchers to perform analysis for the identification of mutations, but also provides detailed information of mutations’ positions, amino acid changes as well as antiviral information and related literature of resistance mutations to the AAD. When BMA is compared with other available tools (e.g., HCV.geno2pheno), it is different because it provides details of the nucleotides changes that produce an amino acid change.\n\nWe obtained an online information system “BMA” that was designed and developed, for performing mutation analysis. BMA provides a suitable analysis facilitating all data management. The results can be visualized in a text report as well as graphically.\n\nBMA provides a quick, easy, and effective computer-based analysis of mutations, including complete documentation and examples. Furthermore, the development of different visualization techniques allows for proper interpretation and understanding of the results. The data obtained by BMA will be useful for the assessment and surveillance of HCV resistance to new antivirals, and for the treatment regimens by selecting those DAAs to which the virus is not resistant, avoiding unnecessary treatment failures.\n\nBMA has been designed to be flexible and adaptable. It is a great advantage because it can be used for future evaluation of other viruses such as Influenza and even microorganisms such as bacteria or parasites. Thus, as future work, BMA will analyze a wide range of pathogens. In addition, BMA might be upgraded in order to offer new visualization techniques for facilitating the interpretation of the obtained analysis.\n\nBMA has a small disadvantage. It requires a specific format of sequence information, which is very similar to the FASTA format; thus, the preparation of such information might require a small additional effort. In addition, in future versions, BMA will accept different file formats such as the FASTA format.\n\n\nSoftware availability\n\nSoftware available from: http://bma.itiud.org\n\nLatest source code: https://github.com/florezfernandez/bma\n\nArchived source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.5099427\n\nLicense: GNU General Public License (GPL)",
"appendix": "Author contributions\n\n\n\nKarina Salvatierra reviewed the literature and wrote the manuscript.\n\nHector Florez designed and developed the BMA software and wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work has been supported by the Information Technologies Innovation (ITI) Research Group.\n\n\nAcknowledgments\n\nThe authors would like to thank the ITI Research Group that belongs to the Universidad Distrital Francisco José de Caldas (Colombia), and Professor Jorge E. Osorio, Department of Pathobiological Sciences, University of Wisconsin-Madison (USA), for its collaboration in the project, making this research possible.\n\n\nReferences\n\nHajarizadeh B, Grebely J, Dore GJ: Epidemiology and natural history of HCV infection. Nat Rev Gastroenterol Hepatol. 2013; 10(9): 553–62. PubMed Abstract | Publisher Full Text\n\nMohd Hanafiah K, Groeger J, Flaxman AD, et al.: Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013; 57(4): 1333–42. PubMed Abstract | Publisher Full Text\n\nAlter MJ: HCV routes of transmission: what goes around comes around. Semin Liver Dis. 2011; 31(4): 340–6. PubMed Abstract | Publisher Full Text\n\nPondé RA: Hidden hazards of HCV transmission. Med Microbiol Immunol. 2011; 200(1): 7–11. PubMed Abstract | Publisher Full Text\n\nSalvatierra K, Florez H: Analysis of hepatitis C virus in hemodialysis patients. Infectio. 2015. Publisher Full Text\n\nMcHutchison JG, Lawitz EJ, Shiffman ML, et al.: Peginterferon alfa-2b or alfa-2a with ribavirin for treatment of hepatitis C infection. N Engl J Med. 2009; 361(6): 580–93. PubMed Abstract | Publisher Full Text\n\nPawlotsky JM: Treatment of chronic hepatitis C: current and future. Curr Top Microbiol Immunol. 2013; 369: 321–42. PubMed Abstract | Publisher Full Text\n\nCortez KJ, Maldarelli F: Clinical management of HIV drug resistance. Viruses. 2011; 3(4): 347–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalvatierra K, Fareleski S, Forcada A, et al.: Hepatitis C virus resistance to new specifically-targeted antiviral therapy: A public health perspective. World J Virol. 2013; 2(1): 6–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKwong AD, McNair L, Jacobson I, et al.: Recent progress in the development of selected hepatitis C virus NS3.4A protease and NS5B polymerase inhibitors. Curr Opin Pharmacol. 2008; 8(5): 522–31. PubMed Abstract | Publisher Full Text\n\nWelsch C, Domingues FS, Susser S, et al.: Molecular basis of telaprevir resistance due to V36 and T54 mutations in the NS3-4A protease of the hepatitis C virus. Genome Biol. 2008; 9(1): R16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWyles DL: Antiviral resistance and the future landscape of hepatitis C virus infection therapy. J Infect Dis. 2013; 207(Suppl 1): S33–9. PubMed Abstract | Publisher Full Text\n\nLópez-Labrador FX, Moya A, Gonzàlez-Candelas F: Mapping natural polymorphisms of hepatitis C virus NS3/4A protease and antiviral resistance to inhibitors in worldwide isolates. Antivir Ther. 2008; 13(4): 481–94. PubMed Abstract\n\nMargeridon S, Le Pogam S, Liu T, et al.: Ultra-deep sequencing of the NS3 and NS5B regions detects pre-existing resistant variants to direct acting antivirals (DAA) in HCV genotype 1 treatment-naive infected patients. Hepatology. 2010; 52(Suppl 1): 714A. Reference Source\n\nMargeridon S, Le Pogam S, Liu TF, et al.: No detection of variants bearing NS5B S282T mericitabine (MCB) resistance mutation in DAA treatment-naive HCV genotype 1 infected patients using ultra-deep pyrosequencing (UDPS). Hepatology. 2011; 54(Suppl 1): 532A. Reference Source\n\nWyles DL, Gutierrez JA: Importance of HCV genotype 1 subtypes for drug resistance and response to therapy. J Viral Hepat. 2014; 21(4): 229–40. PubMed Abstract | Publisher Full Text\n\nSarrazin C, Kieffer TL, Bartels D, et al.: Dynamic hepatitis C virus genotypic and phenotypic changes in patients treated with the protease inhibitor telaprevir. Gastroenterology. 2007; 132(5): 1767–77. PubMed Abstract | Publisher Full Text\n\nSarrazin C, Rouzier R, Wagner F, et al.: SCH 503034, a novel hepatitis C virus protease inhibitor, plus pegylated interferon alpha-2b for genotype 1 nonresponders. Gastroenterology. 2007; 132(4): 1270–8. PubMed Abstract | Publisher Full Text\n\nKuntzen T, Timm J, Berical A, et al.: Naturally occurring dominant resistance mutations to hepatitis C virus protease and polymerase inhibitors in treatment-naïve patients. Hepatology. 2008; 48(6): 1769–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSusser S, Vermehren J, Forestier N, et al.: Analysis of long-term persistence of resistance mutations within the hepatitis C virus NS3 protease after treatment with telaprevir or boceprevir. J Clin Virol. 2011; 52(4): 321–7. PubMed Abstract | Publisher Full Text\n\nThompson AJ, McHutchison JG: Antiviral resistance and specifically targeted therapy for HCV (STAT-C). J Viral Hepat. 2009; 16(6): 377–87. PubMed Abstract | Publisher Full Text\n\nExcoffier L, Heckel G: Computer programs for population genetics data analysis: a survival guide. Nat Rev Genet. 2006; 7(10): 745–58. PubMed Abstract | Publisher Full Text\n\nRumbaugh J, Jacobson I, Booch G: The Unified Modeling Language Reference Manual. Pearson Higher Education. 2004. Reference Source\n\nSalvatierra KA: Resistencias a nuevos antivirales de acción directa en aislados clínicos del virus de la hepatitis C. [Ph.D. thesis]. Valencia. Universitat de Valencia, Spain. 2014. Reference Source\n\nPatiño-Galindo JA, Salvatierra K, González-Candelas F, et al.: Comprehensive Screening for Naturally Occurring Hepatitis C Virus Resistance to Direct-Acting Antivirals in the NS3, NS5A, and NS5B Genes in Worldwide Isolates of Viral Genotypes 1 to 6. Antimicrob Agents Chemother. 2016; 60(4): 2402–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994; 22(22): 4673–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlorez H, Salvatierra K: BMA. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "14898",
"date": "15 Jul 2016",
"name": "Vicente Pérez-Brocal",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article \"Biomedical mutation Analysis (BMA): A software tool for analyzing mutations associated with antiviral resistance\" describes a new online tool intended for detection of mutations in amino acids implicated in resistance to direct-acting antivirals.\nDespite the complexity and high degree of specifications reflected in the manuscript, especially in the material and methods section, the results section describes each step in a straightforward way that facilitates the implementation of this software. The usage of the language makes this article accessible even for non-experts in the field.\nThe methodology has been validated using the nucleotide sequence of some gene from HCV genotype 1b, and a compilation of resistance mutations previously described in the literature for those genes.\nMinor changes:\nIn Materials and Methods section, third paragraph, begins as \" The database was design using the tool\" and should say \"The database was designed using the tool\".\n\nSentence \"For reliable calculations the sequences must contain a substantial part of the genes NS3, NS5A or NS5B\" in the Results section results quite vague. How much is a substantial part?\n\nAAD, that appears in the first paragraph of the conclusions, should be replaced with DAA, for congruence with the rest of the text.",
"responses": [
{
"c_id": "2117",
"date": "02 Aug 2016",
"name": "karina salvatierra",
"role": "Author Response",
"response": "Dear Vicente Perez-Brocal Thank you so much for your feedback. Best Regards"
}
]
},
{
"id": "14899",
"date": "28 Jul 2016",
"name": "Carla García-Morales",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors of the BMA tool propose the creation of a bioinformatics tool that identifies changes in nucleotide and amino acid sequences which could represent an impact on antiviral resistance. The suggested system is a promising idea that could be useful for biomedical and pharmacological areas as there is no freely available similar tools.\nOne of my concerns, if not the main is English vocabulary and grammar, I would suggest the authors to take advantage of an editing service as some statements could be improved. For a start, from the BMA home page, “In biology, a mutation is a change of the nucleotide sequence of the genome of an org…Mutations play a part in both normal and abnormal biological processes including: evolution, cancer, and the development of the immune system, including junction diversity…Mutation can result in several different types of change in sequences. Mutations in genes can have no effect, alter the product of a gene, or prevent the gene from functioning properly or completely.” I would suggest “In biology, a mutation is a change in the nucleotide sequence on the genome of an org…Mutations play an important role in both normal and abnormal biological processes such as: evolution, cancer, and the development of the immune system, including junction diversity. Mutations can result in several different types of sequence changes; such changes could have no effect, alter the product of the gene, or prevent it from functioning properly or completely.”\nIt would also be advisable that authors present a tutorial link in the BMA site.\nAs for the manuscript, authors should explain for figure 2 what is shown under the column named “Main”; and again, be careful about expressions such as scientifically support.\nAs for any publication on bioinformatics tools, F1000 ask the author to provide sufficient details of codes used for the implementation and operation, unfortunately could not find this information; The only code needed to insert different sequences is the > symbol as in other bioinformatics programs, however the authors should be very clear about this, for example, in the 4th line of number 3 in Results, they only mention “BMA can recognize plain text files, but they have to follow a specific format”, so what do they mean with specific format?, this line is not needed as they explain the required format immediately. Another point that should be improved is in Figure 3. Introduce the patient sequence in a text file format instead of “plain text”\nI would also suggest to change figure 6, for one where the pointer is over the node so that the reader will be able to see what each node represents.",
"responses": [
{
"c_id": "2116",
"date": "02 Aug 2016",
"name": "karina salvatierra",
"role": "Author Response",
"response": "Dear Carla Garcia-Morales. Thank you so much for your feedback. We want to inform you: We have updated the text in bma.itiud.org as you suggested. In bma.itiud.org, there is in the menu a link called Screencast. It is a video tutorial that presents how to use BMA Information regarding the list of positions has been updating including an explanation of the column “Main” In the section “Software availability”, we provide all details about the source code of the project. Indeed, the project is open source, so everyone can download it. Figure 6 has been also updated in order to present the popup messages of the force directed graph. In addition, the explanation of this figure has been extended. Thank you so much Best Regards"
}
]
}
] | 1
|
https://f1000research.com/articles/5-1141
|
https://f1000research.com/articles/5-1438/v1
|
20 Jun 16
|
{
"type": "Software Tool Article",
"title": "From reads to genes to pathways: differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline",
"authors": [
"Yunshun Chen",
"Aaron T. L. Lun",
"Gordon K. Smyth",
"Yunshun Chen",
"Aaron T. L. Lun"
],
"abstract": "In recent years, RNA sequencing (RNA-seq) has become a very widely used technology for profiling gene expression. One of the most common aims of RNA-seq profiling is to identify genes or molecular pathways that are differentially expressed (DE) between two or more biological conditions. This article demonstrates a computational workflow for the detection of DE genes and pathways from RNA-seq data by providing a complete analysis of an RNA-seq experiment profiling epithelial cell subsets in the mouse mammary gland. The workflow uses R software packages from the open-source Bioconductor project and covers all steps of the analysis pipeline, including alignment of read sequences, data exploration, differential expression analysis, visualization and pathway analysis. Read alignment and count quantification is conducted using the Rsubread package and the statistical analyses are performed using the edgeR package. The differential expression analysis uses the quasi-likelihood functionality of edgeR.",
"keywords": [
"RNA sequencing",
"molecular pathways",
"gene expression",
"R software"
],
"content": "Introduction\n\nIn recent years, RNA sequencing (RNA-seq) has become a very widely used technology for profiling transcriptional activity in biological systems. One of the most common aims of RNA-seq profiling is to identify genes or molecular pathways that are differentially expressed (DE) between two or more biological conditions. Changes in expression can then be associated with differences in biology, providing avenues for further investigation into potential mechanisms of action.\n\nThis article provides a detailed workflow for analyzing an RNA-seq study from the raw reads through to differential expression and pathway analysis using Bioconductor packages1. The article gives a complete analysis of RNA-seq data that were collected to study the effects of pregnancy and lactation on the luminal cell lineage in the mouse mammary gland2. The pipeline uses the Rsubread package3 for mapping reads and assigning them to genes, and the edgeR package4 for statistical analyses.\n\nRNA-seq analysis involves a number of steps, including read alignment, read summarization, differential expression and pathway analysis. Here we use the Subread aligner3 for mapping and featureCounts5 for assigning reads to genes. As well as being fast and efficient, these algorithms have the advantage of having native implementations as R functions in the Rsubread package. This means that the entire analysis can be conducted efficiently within the R environment.\n\nThe workflow uses edgeR's quasi-likelihood pipeline (edgeR-quasi) for differential expression. This statistical methodology uses negative binomial generalized linear models6 but with F-tests instead of likelihood ratio tests7. This method provides stricter error rate control than other negative binomial based pipelines, including the traditional edgeR pipelines6,8,9 or DESeq210. The edgeR-quasi pipeline is based on similar statistical methodology to that of the QuasiSeq package7, which has performed well in third-party comparisons11. Compared to QuasiSeq, the edgeR functions offer speed improvements and some additional statistical refinements12. The RNA-seq pipelines of the limma package also offer excellent error rate control13,14. While the limma pipelines are recommended for large-scale datasets, because of their speed and flexibility, the edgeR-quasi pipeline gives better performance in low-count situations15,16. For the data analyzed here, the edgeR-quasi, limma-voom and limma-trend pipelines are all equally suitable and give similar results.\n\nThe analysis approach illustrated in this article can be applied to any RNA-seq study that includes some replication, but it is especially appropriate for designed experiments with multiple treatment factors and with small numbers of biological replicates. The approach assumes that RNA samples have been extracted from cells of interest under two or more treatment conditions, that RNA-seq profiling has been applied to each RNA sample and that there are independent biological replicates for at least one of the treatment conditions. The Rsubread part of the workflow takes FASTQ files of raw sequence reads as input, while the edgeR part of the pipeline takes a matrix of genewise read counts as input.\n\n\nDescription of the biological experiment\n\nThis workflow demonstrates a complete bioinformatics analysis of an RNA-seq study that is available from the GEO repository as series GSE60450. The RNA-seq data were collected to study the lineage of luminal cells in the mouse mammary gland and in particular how the expression profiles of the members of the lineage change upon pregnancy and lactation2. Specifically, the study examined the expression profiles of basal stem-cell enriched cells (B) and committed luminal cells (L) in the mammary glands of virgin, pregnant and lactating mice. There are therefore six groups of RNA samples, one for each combination of cell type and mouse status. Two biological replicates were collected for each group.\n\nThis study used an Illumina Hiseq sequencer to generate about 30 million 100bp single-end reads for each sample. Subread version 1.4.4 (http://subread.sourceforge.net) was used to align the reads to the mouse mm10 genome and featureCounts was used to assign reads to Entrez Genes using RefSeq gene annotation. The FASTQ files containing the raw sequence reads were deposited to the Sequence Read Archive (SRA) repository and the read counts were deposited to GEO.\n\nThis experimental design is summarized in the table below, where the basal and luminal cell types are abbreviated with B and L respectively. The GEO and SRA identifiers for each RNA sample are also shown:\n\n\n\nThe experiment can be viewed as a one-way layout with six groups. For later use, we combine the treatment factors into a single grouping factor:\n\n\n\n\nPreliminary analysis\n\nReaders wishing to reproduce the analysis presented in this article can either download the matrix of read counts from GEO or recreate the read count matrix from the raw sequence counts. We will present first the analysis using the downloaded matrix of counts. At the end of this article we will present the R commands needed to recreate this matrix.\n\nThe following commands download the genewise read counts for the GEO series GSE60450. The zipped tab-delimited text file GSE60450_Lactation-GenewiseCounts.txt.gz will be downloaded to the working R directory:\n\nThe counts can then read into a data.frame in R:\n\nThe row names of GenewiseCounts are the Entrez Gene Identifiers. The first column contains the length of each gene, being the total number of bases in exons or UTRs for that gene. The remaining 12 columns contain read counts and correspond to rows of targets.\n\nThe edgeR package stores data in a simple list-based data object called a DGEList. This object is easy to use as it can be manipulated like an ordinary list in R, and it can also be subsetted like a matrix. The main components of a DGEList object are a matrix of read counts, sample information in the data.frame format and optional gene annotation. We enter the counts into a DGEList object using the function DGEList in edgeR:\n\nThe Entrez Gene Ids link to gene information in the NCBI database. The org.Mm.eg.db package can be used to complement the gene annotation information. Here, a column of gene symbols is added to y$genes:\n\nEntrez Ids that no longer have official gene symbols are dropped from the analysis. The whole DGEList object, including annotation as well as counts, can be subsetted by rows as if it was a matrix:\n\nGenes that have with very low counts across all the libraries should be removed prior to downstream analysis. This is justified on both biological and statistical grounds. From biological point of view, a gene must be expressed at some minimal level before it is likely to be translated into a protein or to be considered biologically important. From a statistical point of view, genes with consistently low counts are very unlikely be assessed as significantly DE because low counts do not provide enough statistical evidence for a reliable judgement to be made. Such genes can therefore be removed from the analysis without any loss of information.\n\nThe downstream differential expression analysis is not sensitive to the exact number of genes that are filtered. As a rule of thumb, we require that gene have a count of at least 10–15 in at least some libraries before it is considered to be expressed in the study. To account for differences in library sizes between samples, the filtering process is based on the count-per-million (CPM) values rather than on the counts directly.\n\nFor the current analysis, we keep genes that have CPM values above 0.5 in at least two libraries:\n\nA CPM of 0.5 is used as it is equivalent to a count of 12–14 for the library sizes in this data set. We demand that a gene is expressed in at least two libraries because each group contains two replicates. This ensures that a gene will be retained if it is expressed in all the libraries belonging to any of the six groups.\n\nThe DGEList object is subsetted to retain only the non-filtered genes:\n\nNote that keep.lib.sizes=FALSE causes the library sizes to be recomputed after the filtering. This is generally recommended, although the effect on the downstream analysis is usually small.\n\nNote that the filtering rule should not make any reference to which RNA libraries belong to which group, because doing so would bias the subsequent differential expression analysis.\n\nNormalization by trimmed mean of M values (TMM)17 is performed by using the calcNormFactors function, which returns the DGEList argument with only the norm.factors changed. It calculates a set of normalization factors, one for each sample, to eliminate composition biases between libraries. The product of these factors and the library sizes defines the effective library size, which replaces the original library size in all downstream analyses.\n\n\n\nThe normalization factors of all the libraries multiply to unity. A normalization factor below one indicates that a small number of high count genes are monopolizing the sequencing, causing the counts for other genes to be lower than would be usual given the library size. As a result, the library size will be scaled down, analogous to scaling the counts upwards in that library. Conversely, a factor above one scales up the library size, analogous to downscaling the counts.\n\nThe performance of the TMM normalization can be assessed by mean-difference (MD) plots. This visualizes the library size-adjusted log-fold change between two libraries (the difference) against the average log-expression across those libraries (the mean). The following command produces an MD plot which compares sample 1 to an artificial reference library constructed from the average of all other samples:\n\n\n\n(see Figure 1). The bulk of the genes should be centered at a line of zero log-fold change if the composition bias between libraries has been removed successfully. The quality check should be repeated with an MD plot for each of the other samples.\n\nEach point represents a gene, and the red line indicates a log-ratio of zero. The majority of points cluster around the red line.\n\nThe RNA samples can be clustered in two dimensions using multi-dimensional scaling (MDS) plots. This is a quality control step to explore the overall differences between the expression profiles of the different samples. Here we decorate the MDS plot to indicate the cell groups:\n\n(see Figure 2). In the MDS plot, the distance between each pair of samples can be interpreted as the leading log-fold change between the samples for the genes that best distinguish that pair of samples. By default, leading fold-change is defined as the root-mean-square of the largest 500 log2-fold changes between that pair of samples. Figure 2 shows that replicate samples from the same group cluster together while samples from different groups are well separated. In other words, differences between groups are much larger than those within groups, meaning that there are likely to be statistically significant differences between the groups. The distance between basal cells on the left and luminal cells on the right is about six units on the x-axis, corresponding to a leading fold change of about 64-fold between the two cell types. The differences between the virgin, pregnant and lactating expression profiles appear to be magnified in luminal cells compared to basal.\n\nSamples are separated by the cell type in the first dimension, and by the mouse status in the second dimension.\n\nLinear modeling and differential expression analysis in edgeR requires a design matrix to be specified. The design matrix records which treatment conditions were applied to each samples, and it also defines how the experimental effects are parametrized in the linear models. The experimental design for this study can be viewed as a one-way layout and the design matrix can be constructed in a simple and intuitive way by:\n\nThis design matrix simply links each group to the samples that belong to it. Each row of the design matrix corresponds to a sample whereas each column represents a coefficient corresponding to one of the six groups.\n\nedgeR uses the negative binomial (NB) distribution to model the read counts for each gene in each sample. The dispersion parameter of the NB distribution accounts for variability between biological replicates6. edgeR estimates an empirical Bayes moderated dispersion for each individual gene. It also estimates a common dispersion, which is a global dispersion estimate averaged over all genes, and a trended dispersion where the dispersion of a gene is predicted from its abundance. Dispersion estimates are most easily obtained from the estimateDisp function:\n\nThis returns a DGEList object with additional components added to hold the estimated dispersions. Here robust=TRUE has been used to protect the empirical Bayes estimates against the possibility of outlier genes with exceptionally large or small individual dispersions18.\n\nThe dispersion estimates can be visualized with plotBCV:\n\n\n\n(see Figure 3). The vertical axis of the plotBCV plot shows square-root dispersion, also known as biological coefficient of variation (BCV)6.\n\nThe plot shows the square-root estimates of the common, trended and tagwise NB dispersions.\n\nFor RNA-seq studies, the NB dispersions tend to be higher for genes with very low counts. The dispersion trend tends to decrease smoothly with abundance and to asymptotic to a constant value for genes with larger counts. From our past experience, the asymptotic value for the BCV tends to be in range from 0.05 to 0.2 for genetically identical mice or cell lines, whereas somewhat larger values (> 0.3) are observed for human subjects.\n\nThe NB model can be extended with quasi-likelihood (QL) methods to account for gene-specific variability from both biological and technical sources7,12. Under the QL framework, the NB dispersion trend is used to describe the overall biological variability across all genes, and gene-specific variability above and below the overall level is picked up by the QL dispersion. In the QL approach, the tagwise NB dispersions are not used. The estimation of QL dispersions is performed using the glmQLFit function:\n\n\n\nThis returns a DGEGLM object with the estimated values of the GLM coefficients for each gene. It also contains a number of empirical Bayes (EB) statistics including the QL dispersion trend, the squeezed QL dispersion estimates and the prior degrees of freedom (df). The QL dispersions can be visualized by plotQLDisp:\n\n\n\n(see Figure 4).\n\nEstimates are shown for the raw (before EB moderation), trended and squeezed (after EB moderation) dispersions. Note that the QL dispersions and trend shown here are relative to the NB dispersion trend show in Figure effig:plotBCV.\n\nThe QL functions moderate the genewise the QL dispersion estimates in the same way that the limma package moderates variances19. The raw QL dispersion estimates are squeezed towards a global trend, and this moderation reduces the uncertainty of the estimates and improves testing power. The extent of the squeezing is governed by the value of the prior df estimated from the data. Large prior df estimates indicate that the QL dispersions are less variable between genes, meaning that strong EB moderation should be performed. Smaller prior df estimates indicate that the true unknown dispersions are highly variable, so weaker moderation towards the trend is appropriate.\n\n\n\nSetting robust=TRUE in glmQLFit is usually recommended18. This allows gene-specific prior df estimates, with lower values for outlier genes and higher values for the main body of genes. This reduces the chance of getting false positives from genes with extremely high or low raw dispersions, while at the same time increasing statistical power to detect differential expression for the main body of genes.\n\n\nDifferential expression analysis\n\nThe next step is to test for differential expression between the experimental groups. One of the most interesting comparisons is that between the basal pregnant and lactating groups. The contrast corresponding to any specified comparison can be constructed conveniently using the makeContrasts function:\n\nIn subsequent results, a positive log2-fold-change (logFC) will indicate a gene up-regulated in lactating mice relative to pregnant, whereas a negative logFC will indicate a gene more highly expressed in pregnant mice. We will use QL F-tests instead of the more usual likelihood ratio tests (LRT) as they give stricter error rate control by accounting for the uncertainty in dispersion estimation:\n\nThe top DE genes can be viewed with topTags:\n\n\n\nIn order to control the false discovery rate (FDR), multiple testing correction is performed using the Benjamini-Hochberg method. The top DE gene Csn1s2b has a large positive logFC, showing that it is far more highly expressed in the basal cells of lactating than pregnant mice. This gene is indeed known to be a major source of protein in milk.\n\nThe total number of DE genes identified at an FDR of 5% can be shown with decideTestsDGE. There are in fact more than 5000 DE genes in this comparison:\n\n\n\nThe magnitude of the differential expression changes can be visualized with a fitted model MD plot:\n\n\n\n(see Figure 5). The logFC for each gene is plotted against the average abundance in log2-CPM, i.e., logCPM in the table above. Genes that are significantly DE are highlighted:\n\nSignificantly up and down DE genes are highlighted in red and blue, respectively.\n\nglmQLFTest identifies differential expression based on statistical significance regardless of how small the difference might be. For some purposes we might be interested only in genes with reasonably large expression changes. The above analysis found more than 5000 DE genes between the basal pregnant and lactating groups. With such a large number of DE genes, it makes sense to narrow down the list to genes that are more biologically meaningful.\n\nA commonly used approach is to apply FDR and logFC cutoffs simultaneously. However this tends to favor lowly expressed genes, and also fails to control the FDR correctly. A better and more rigorous approach is to modify the statistical test so as to detect expression changes greater than a specified threshold. In edgeR, this can be done using the glmTreat function. This function is analogous to the TREAT method for microarrays20 but is adapted to the NB framework. Here we test whether the differential expression fold changes are significantly greater than 1.5, that is, whether the logFCs are significantly greater than log2(1.5):\n\n\n\nNote that the argument lfc is an abbreviation for “log-fold-change”. About 1100 genes are detected as DE with a FC significantly above 1.5 at an FDR cut-off of 5%.\n\n\n\nThe test results can be visualized in an MD plot:\n\n\n\n(see Figure 6).\n\nGenes with fold-changes significantly greater than 1.5 are highlighted.\n\nThe p-values obtained by glmTreat are usually larger than those from glmQLFTest because the latter is testing the null hypothesis that the true logFC is zero. glmTreat is testing a different hypothesis and requires stronger evidence for differential expression than conventional tests do. It provides greater specificity for identifying the most important genes with large fold changes.\n\nNote that the logFC threshold in glmTreat is not the same as a logFC cutoff. Genes will need to exceed this threshold by some way before being declared statistically significant. It is better to interpret the threshold as the FC below which we are definitely not interested in the gene rather than the FC above which we are interested in the gene. The value of the FC threshold can be varied depending on the dataset. In the presence of a huge number of DE genes, a relatively large FC threshold may be appropriate to narrow down the search to genes of interest. In the absence of DE genes, on the other hand, a small or even no FC threshold shall be used. If the threshold level is set to zero, then glmTreat reduces to glmQLFTest depending on the pipeline used in the analysis. glmTreat can also be used with other edgeR pipelines, although we don’t show that in this workflow.\n\nHeatmaps are a popular way to display differential expression results for publication purposes. To create a heatmap, we first convert the read counts into log2-counts-per-million (logCPM) values. This can be done with the cpm function:\n\n\n\nThe introduction of prior.count is to avoid undefined values and to reduce the variability of the logCPM values for genes with low counts. Larger values for prior.count shrink the logFCs for low count genes towards zero.\n\nWe will create a heatmap to visualize the top 30 DE genes according to the TREAT test between B.lactating and B.pregnant. The advantage of a heatmap is that it can display the expression pattern of the genes across all the samples. First we select the logCPM values for the 30 top genes:\n\n\n\nThen we scale each row (each gene) to have mean zero and standard deviation one:\n\n\n\nA heat map can then be produced by the heatmap.2 function in the gplots package:\n\n\n\n(see Figure 7). By default, heatmap.2 clusters genes and samples based on Euclidean distance between the expression values. Considering that we have pre-standardized the rows of the logCPM matrix, the use of Euclidean distance for the standardize values is equivalent to Pearson correlation between genes for the original logCPM values. As expected, samples from the same group are clustered together.\n\nThe differential expression analysis comparing two groups can be easily extended to comparisons between three or more groups. This is done by creating a matrix of independent contrasts. In this manner, users can perform a one-way analysis of deviance (ANODEV) for each gene.\n\nSuppose we want to compare the three groups in the luminal population, i.e., virgin, pregnant and lactating. An appropriate contrast matrix can be created as shown below, to make pairwise comparisons between all three groups:\n\nThe QL F-test is then applied to identify genes that are DE between the three groups. This combines the three pairwise comparisons into a single F-statistic and p-value. The top set of significant genes can be displayed with topTags:\n\n\n\nNote that the three contrasts of pairwise comparisons are linearly dependent. Constructing the contrast matrix with any two of the contrasts would be sufficient for an ANODEV test. If the contrast matrix contains all three possible pairwise comparisons, then only the log-fold changes of the first two contrasts are shown in the output of topTags.\n\nThe flexibility of the GLM framework makes it possible to specify arbitrary contrasts for differential expression tests. Suppose we are interested in testing whether the change in expression between lactating and pregnant mice is the same for basal cells as it is for luminal cells. In statistical terminology, this is the interaction effect between mouse status and cell type. The contrast corresponding to this testing hypothesis can be made as follows.\n\n\n\nThen the QL F-test is conducted to identify genes that are DE under this contrast. The top set of DE genes are viewed with topTags.\n\n\n\n\nPathway analysis\n\nWe now consider the problem of interpreting the differential expression results in terms of higher order biological processes or molecular pathways. One of the most common used resources is gene ontology (GO) databases, which annotate genes according to a dictionary of annotation terms. A simple and often effective way to interpret the list of DE genes is to count the number of DE genes that are annotated with each possible GO term. GO terms that occur frequently in the list of DE genes are said to be over-represented or enriched. In edgeR, GO analyses can be conveniently conducted using the goana function. Here were apply goana to the output of the TREAT analysis comparing B.lactating to B.pregant. The top most significantly enriched GO terms can be viewed with topGO.\n\n\n\nThe goana function automatically extracts DE genes from the tr object, and conducts overlap tests for the up and down DE genes separately. The row names of the output are the universal identifiers of the GO terms and the Term column gives the human-readable names of the terms. The Ont column shows the ontology domain that each GO term belongs to. The three domains are: biological process (BP), cellular component (CC) and molecular function (MF). The N column represents the total number of genes annotated with each GO term. The Up and Down columns indicate the number of genes within the GO term that are significantly up- and down-regulated in this differential expression comparison, respectively. The P.Up and P.Down columns contain the p-values for over-representation of the GO term in the up- and down-regulated genes, respectively. By default the output table from topGO is sorted by the minimum of P.Up and P.Down. Other options are available. For example, topGO(go, sort=\"up\") lists the top GO terms that are over-represented in the up-regulated genes. The domain of the enriched GO terms can also be specified by users. For example, topGO(go, ontology=\"BP\") returns the top GO terms belonging to the biological process domain. This avoids other domains that are not of interest.\n\nThe goana function uses the NCBI RefSeq annotation and requires the use of Entrez Gene Identifiers.\n\nAnother popular annotation database is the Kyoto Encyclopedia of Genes and Genomes (KEGG). Much smaller than GO, this is a curated database of molecular pathways and disease signatures. A KEGG analysis can be done exactly as for GO, but using the kegga function:\n\n\n\nThe output from topKEGG is the same as from topGO except that row names become KEGG pathway IDs, Term becomes Pathway and there is no Ont column. Both the GO and KEGG analyses show that the cell cycle pathway is strongly down-regulated upon lactation in mammary stem cells.\n\nBy default, the kegga function automatically reads the latest KEGG annotation from the Internet each time it is run. The KEGG database uses Entrez Gene Ids, and the kegga function assumes these are available as the row names of tr.\n\nThe GO and KEGG analyses shown above are relatively simple analyses that rely on a list of DE genes. The list of DE genes is overlapped with the various GO and KEGG annotation terms. The results will depend on the significance threshold that is used to assess differential expression.\n\nIf the aim is to test for particular gene expression signatures or particular pathways, a more nuanced approach is to conduct a roast or fry gene set test21. These functions test whether a set of genes is DE, assessing the whole set of genes as a whole. Gene set tests consider all the genes in the specified set and do not depend on any pre-emptive significance cutoff. The set of genes can be chosen to be representative of any pathway or phenotype of interest.\n\nroast gives p-values using random rotations of the residual space. In the edgeR context, fry is generally recommended over roast. fry gives an accurate analytic approximation to the results that roast would give, with default settings, if an extremely large number of rotations was used.\n\nHere, suppose we are interested in three GO terms related to cytokinesis. Each GO term is used to define a set of genes annotated with that term. The names of these terms are shown below:\n\nThe first step is to extract the genes associated with each GO term from the GO database. This produces a list of three components, one for each GO term. Each component is a vector of Entrez Gene IDs for that GO term:\n\n\n\nSuppose the comparison of interest is between the virgin and lactating groups in the basal population. We can use fry to test whether the cytokinesis GO terms are DE for this comparison:\n\n\n\nEach row of the output corresponds to a gene set. The NGenes column provides the number of genes in each set. The Direction column indicates the net direction of change. The PValue column gives the two-sided p-value for testing whether the set in DE as a whole, either up or down. The PValue.Mixed column gives a p-value for testing whether genes in the set tend to be DE, without regard to direction. The PValue column is appropriate when genes in the set are expected to be co-regulated, all or most changing expression in the same direction. The PValue.Mixed column is appropriate when genes in the set are not necessarily co-regulated or may be regulated in different directions for the contrast in question. FDRs are calculated from the corresponding p-values across all sets.\n\nThe results of a gene set test can be viewed in a barcode plot produced by the barcodeplot function. Suppose visualization is performed for the gene set defined by the GO term GO:0032465:\n\n(see Figure 8). In the plot, all genes are ranked from left to right by decreasing log-fold change for the contrast and the genes within the gene set are represented by vertical bars, forming the barcode-like pattern. The curve (or worm) above the barcode shows the relative local enrichment of the bars in each part of the plot. The dotted horizontal line indicates neutral enrichment; the worm above the dotted line shows enrichment while the worm below the dotted line shows depletion. In this particular barcode plot the worm shows enrichment on the left for positive logFCs, and depletion on the right for negative logFCs. The conclusion is that genes associated with this GO term tend to be up-regulated in the basal cells of virgin mice compared to lactating mice, confirming the result of the fry test above.\n\nX-axis shows logFC for B.virgin vs B.lactating. Black bars represent genes annotated with the GO term. The worm shows relative enrichment.\n\nFinally we demonstrate a gene set enrichment style analysis using the Molecular Signatures Database (MSigDB)22. We will use the C2 collection of the MSigDB, which is a collection of nearly 5000 curated gene sets, each representing the molecular signature of a particular biological process or phenotype. The MSigDB itself is purely human, but the Walter and Eliza Hall Institute maintains a mouse version of the database. We load the mouse version of the C2 collection from the WEHI website:\n\nThis will load Mm.c2, which is a list of gene sets, each a vector of Entrez Ids. This can be converted to a list of index numbers:\n\nFirst we compare virgin stem cells to virgin luminal cells:\n\nAs expected, the mammary stem cell and mammary luminal cell signatures from Lim et al23 are top-ranked, and in the expected directions.\n\nWe can visualize the top signature, combining the up and down mammary stem cell signatures to make a bi-directional signature set:\n\n\n\n(see Figure 9).\n\n\nPackages used\n\nThis workflow depends on various packages from version 3.3 of the Bioconductor project, running on R version 3.3.0 or higher. The complete list of the packages used for this workflow are shown below:\n\nRed bars show up signature genes, blue bars show down genes. The worms show relative enrichment.\n\n\n\n\nRead alignment and quantification\n\nWe now revisit the question of recreating the matrix of read counts from the raw sequence reads. Unlike the above workflow, which works for any version of R, read alignment requires Unix or Mac OS and, in practice, a high performance Unix server is recommended. The following tasks require only one Bioconductor package, Rsubread. However the fastq-dump utility from the SRA Toolkit is also required to convert from SRA to FASTQ format. This can be downloaded from the NCBI website (http://www.ncbi.nlm.nih.gov/Traces/sra/?view=software) and installed on any Unix system.\n\nThe first task is to download the raw sequence files, which are stored in SRA format on the SRA repository. The files can be conveniently located and downloaded by visiting the web page for the GEO data series GSE60450 at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE60450, then following the ftp link at the foot of the page.\n\nOnce downloaded to the working directory, the 12 SRA files need to be unpacked into FASTQ format using the fastq-dump utility. The following R code makes a system call to fastq-dump to convert each SRA file:\n\n\n\nThis will produce 12 FASTQ files with following names:\n\n\n\nBefore the sequence reads can be aligned, we need to build an index for the GRCm38/mm10 (Dec 2011) build of the mouse genome. Most laboratories that use Rsubread regularly will already have an index file prepared, as this is a once-off operation for each genome release. If you are using Rsubread for mouse for the first time, then the latest mouse genome build can be downloaded from the NCBI location ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCA_000001635.6_GRCm38.p4/GCA_000001635.6_GRCm38.p4_genomic.fna.gz. (Note that this link is for patch 4 of mm10, which is valid at the time of writing in May 2016. The link will change as new patches are released periodically.) An index can then be built by:\n\n\n\nThe FASTQ files can now be aligned to the mouse genome using the align function:\n\n\n\nThis produces a set of BAM files containing the read alignments for each RNA library. The mapping proportions can be summarized by the propmapped function:\n\n\n\nIdeally, the proportion of mapped reads should be above 80%. By default, only reads with unique mapping locations are reported by Rsubread as being successfully mapped. Restricting to uniquely mapped reads is recommended, as it avoids spurious signal from non-uniquely mapped reads derived from, e.g., repeat regions.\n\nThe read counts for each gene can be quantified using the featureCounts function in Rsubread. Conveniently, the Rsubread package includes inbuilt NCBI RefSeq annotation of the mouse and human genomes. featureCounts generates a matrix of read counts for each gene in each sample:\n\n\n\nThe output is a simple list, containing the matrix of counts (counts), a data frame of gene characteristics (annotation), a vector of file names (targets) and summary mapping statistics (stat):\n\n\n\nThe row names of fc$counts are the Entrez gene identifiers for each gene. The column names are the output file names from align, which we simplify here for brevity:\n\n\n\nThe first six rows of the counts matrix are shown below.\n\n\n\nFinally, a DGEList object can be assembled by:\n\n\n\n\nData and software availability\n\nExcept for the targets file targets.txt, all data analyzed in the workflow is read automatically from public websites as part of the code. All software used is publicly available as part of Bioconductor 3.3, except for the fastq-dump utility, which can be downloaded from NCBI website as described in the text. The article includes the complete code necessary to reproduce the analyses shown. The code will also be made available as an executable Bioconductor workflow at http://www.bioconductor.org/help/workflows.",
"appendix": "Author contributions\n\n\n\nAll authors developed and tested the code workflow. All authors wrote the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the National Health and Medical Research Council (Fellowship 1058892 and Program 1054618 to G.K.S, Independent Research Institutes Infrastructure Support to the Walter and Eliza Hall Institute) and by a Victorian State Government Operational Infrastructure Support Grant.\n\n\nAcknowledgments\n\nThe authors thank Wei Shi and Yang Liao for advice with Rsubread and Yifang Hu for creating the mouse version of the MSigDB.\n\n\nReferences\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFu NY, Rios AC, Pal B, et al.: EGF-mediated induction of Mcl-1 at the switch to lactation is essential for alveolar cell survival. Nat Cell Biol. 2015; 17(4): 365–375. PubMed Abstract | Publisher Full Text\n\nLiao Y, Smyth GK, Shi W: The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 2013; 41(10): e108. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014; 30(7): 923–930. PubMed Abstract | Publisher Full Text\n\nMcCarthy DJ, Chen Y, Smyth GK: Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 2012; 40(10): 4288–4297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLund SP, Nettleton D, McCarthy DJ, et al.: Detecting differential expression in RNA-sequence data using quasi-likelihood with shrunken dispersion estimates. Stat Appl Genet Mol Biol. 2012; 11(5): Article 8, pii: /j/sagmb.2012.11.issue-5/1544-6115.1826/1544-6115.1826.xml. PubMed Abstract | Publisher Full Text\n\nRobinson MD, Smyth GK: Small-sample estimation of negative binomial dispersion, with applications to SAGE data. Biostatistics. 2008; 9(2): 321–332. PubMed Abstract | Publisher Full Text\n\nRobinson MD, Smyth GK: Moderated statistical tests for assessing differences in tag abundance. Bioinformatics. 2007; 23(21): 2881–2887. PubMed Abstract | Publisher Full Text\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurden CJ, Qureshi SE, Wilson SR: Error estimates for the analysis of differential expression from RNA-seq count data. PeerJ. 2014; 2: e576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Chen Y, Smyth GK: It’s DE-licious: A Recipe for Differential Expression Analyses of RNA-seq Experiments Using Quasi-Likelihood Methods in edgeR. Methods Mol Biol. 2016; 1418: 391–416. PubMed Abstract | Publisher Full Text\n\nLaw CW, Chen Y, Shi W, et al.: voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014; 15(2): R29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Smyth GK: De novo detection of differentially bound regions for ChIP-seq data using peaks and windows: controlling error rates correctly. Nucleic Acids Res. 2014; 42(11): e95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Smyth GK: diffHic: a Bioconductor package to detect differential genomic interactions in Hi-C data. BMC Bioinformatics. 2015; 16: 258. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, Oshlack A: A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010; 11(3): R25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhipson B, Lee S, Majewski IJ, et al.: Robust hyperparameter estimation protects against hypervariable genes and improves power to detect differential expression. Ann Appl Stat. 2016; 10. Reference Source\n\nSmyth GK: Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004; 3(1): Article 3. PubMed Abstract | Publisher Full Text\n\nMcCarthy DJ, Smyth GK: Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics. 2009; 25(6): 765–771. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu D, Lim E, Vaillant F, et al.: ROAST: rotation gene set tests for complex microarray experiments. Bioinformatics. 2010; 26(17): 2176–2182. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545–15550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLim E, Wu D, Pal B, et al.: Transcriptome analyses of mouse and human mammary cell subpopulations reveal multiple conserved genes and pathways. Breast Cancer Res. 2010; 12: R21. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "14473",
"date": "29 Jun 2016",
"name": "Conrad J. Burden",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere are two main contenders for off-the-shelf software packages for detecting differential expression from RNA-Seq count data: edge R and DESeq2, both of which model over-dispersed count data with a negative binomial distribution. A complete work flow built around DESeq2 was published recently 1, and this is the analogous complete work flow, starting from raw sequence counts, for edgeR. The example given goes further in that it also includes an analysis right through to a molecular pathway analysis of the most highly differentially expressed genes.\n\nThe work flow is built around a number of edgeR functions which have been developed and improved over years. The most recent development is the inclusion of a method employed in an earlier package, called QuasiSeq, which combines a quasi-likelihood approach to estimating estimating over-dispersion with edgeR’s traditional approach of sharing information across genes. In my own review of packages designed for profiling differential expression from count data 2 (cited as ref. 11 in this paper) I observed using synthetic data that QuasiSeq easily outperformed the then existing available packages in terms of accuracy of claimed p-values and false discovery rates. However, it did have the disadvantage of very poor performance in terms of speed. The functions glmQLFit() and glmQLFtest() in the current work flow perform the quasi-likelihood method, but have overcome the speed performance problem completely and run very rapidly. Incidentally I can confirm that I and my co-authors of 2 have no particular connection with either the edgeR or DESeq groups, or the developers of QuasiSeq.\n\nOne problem I have with the example used in the workflow is that there are only two biological replicates in each condition. In ref. 2 we observed using synthetic data that to get consistently accurate estimates of p-values and false discovery rates it is best to use at least 3 and preferably 4 biological replicates in each condition, even with QuasiSeq. See also the recent review by Schurch et al. 3 whose analysis recommends far more than two biological replicates in general with various software packages available at the time of their analysis. A similar independent analysis of the number of biological replicates recommended for the latest edgeR work flow would be welcome.\n\nI have worked through the example given in the paper, starting with “Downloading the read counts” on page 4, working through to pathway analysis ending on page 20. I have not worked through the “Read alignment and quantification” section starting on page 22. In general I found the work flow easy to follow and informative. I have made the following observations:\n\nWhen I got to the line\n\ny <- DGEList(…, group=group, …)\non page 5, the parameter ‘group’ had not been set. I had to do a bit of detective work and, as a workaround, set it up using the following lines of code:\n\nCellType <- c(rep(\"B\", 6), rep(\"L\", 6)) Status <- rep(c(rep(\"virgin\", 2), rep(\"pregnant\", 2), rep(\"lactating\", 2)),2) group <- paste(CellType, Status, sep=\".\") group <- factor(group)\nThis should be fixed.\n\nSuggestion: For ease of use, could calculating the TMM normalisation factors be built into the function DGEList()? If culling the low-count genes makes a noticeable difference, perhaps this could be done just as easily to the original data frame of counts before applying DGEList().\n\nRegarding the diagnostic plot Figure 1, can it happen that the TMM normalisation doesn’t give an MD plot which is symmetric about zero? And if it does, is there a fix?\n\nAs a future enhancement, could a more user-friendly version of the differential expression analysis be made with estimateDisp(), glmQLFit() and glmQLFtest() all built into a single function? The point is that the job of the first two functions, i.e. calculating the trend dispersion and GLM coefficients, can’t be avoided anyway if you are a biologist wanting to do a differential expression analysis. For many users who are not familiar with the negative binomial model, the diagnostic plots of the BCV (Figure 3) and QL dispersion (Figure 4) are likely to be too arcane to be helpful. In fact the tagwise dispersions in Figure 3 are not actually used by the QL method.\n\nA couple of trivial typos:\n\nPage 5, 4th last line: “Genes that have with very low counts …”, remove “with”.\nCaption to Figure 4: “trend show in Figure effig:plotBCV” should be “trend shown in Fig.3”\nPage 16 last line: “B.pregant” should be B.pregnant”.",
"responses": [
{
"c_id": "2100",
"date": "02 Aug 2016",
"name": "Gordon Smyth",
"role": "Author Response",
"response": "Dear Conrad, Thank you for your thoughtful review and for your positive remarks about the performance of the quasi-likelihood pipeline. We understand that the code as given in the article would not run for you completely without the 'targets.txt' file that is read in the first code line. The whole LaTeX article and all the results are actually generated from a knitr Rnw file. The code and associated files have been submitted to Bioconductor as a workflow but are not yet available from the Bioconductor website. In the meantime, we have made the code and associated files available from our own website at http://bioinf.wehi.edu.au/edgeR/F1000Research2016. We find TMM normalization works well for almost all regular gene expression studies. Different normalization methods are more appropriate for other technologies that yield a lot of zeros, for example single-cell RNA-seq, CRISPR, ChIP-seq and Hi-C. Discussion of those is beyond the scope of this article but we have added a couple of references. Note that the MD plot in Figure 1 does not need to be symmetric, as long as the majority of points cluster around the line, and the article now clarifies this point. Our preference is to provide a modular pipeline, encouraging analysts to examine the results at each step. For example, we can't decide whether filtering or TMM normalization is appropriate for a particular study at the time of forming the DGEList. We also like to encourage users to examine the BCV plot. We find this an informative diagnostic plot even as part of limma pipelines that do use the estimateDisp() results. On the other hand, if the number of samples was very large, one might choose to compute just the NB dispersion trend and not the tagwise values. Thanks for pointing out the typos. They will be fixed in the next revision. Regarding sample sizes, the minimum appropriate sample size depends very much on the context. For example, n=2 may be sufficient when comparing well sorted cell types from genetically identical mice whereas n=10 may not be not nearly enough when comparing whole blood for diseased vs normal patients. The repeatability of the results in our workflow is demonstrated by, among other things, the strong correlation between the current results and earlier microarray results on similar cell populations (Figure 9). For cutting edge biomedical experiments, RNA samples can be very difficult to obtain. While larger sample sizes are always preferable, our philosophy is to perform the best possible data analysis for any experiment that our colleagues believe is scientifically worthwhile. It is our aim that edgeR-QL and limma should give statistically correct results for any sample size, even down to n=2 vs n=1. On this topic, we note that the current edgeR-QL code is more robust than the original QuasiSeq method when the sample sizes are very small. Note that QuasiSeq was based on our best understanding of the mathematics at the time of Lund et al (2012), but some important refinements have been added to the edgeR version since. Here is a very small simulated example with n=2 vs n=1 and no true differential expression. Here QuasiSeq gives FDR values as small as 0 or 0.01, whereas the smallest FDR from glmQLFTest is 0.97: > y <- matrix(rpois(10000*3,lambda=10),10000,3) > library(QuasiSeq) > design0 <- matrix(1,3,1) > design1 <- cbind(1,c(0,1,1)) > design.list<-vector(\"list\",2) > design.list[[1]] <- design1 > design.list[[2]] <- design0 > fit <- QL.fit(y, design.list) > res <- QL.results(fit) > lapply(res$Q.values, min) $QL [1] 0 $QLShrink [1] 0.0138122 $QLSpline [1] 0.01351794 > library(edgeR) Loading required package: limma > dge <- DGEList(counts=y) > dge <- estimateDisp(dge, design1) > fit <- glmQLFit(dge, design1) > ql <- glmQLFTest(fit) > topTags(ql) Coefficient: logFC logCPM F PValue FDR 7645 -1.911573 6.981129 42.64028 0.008218779 0.9733061 4209 3.366857 6.640090 41.30839 0.008583583 0.9733061 4669 2.451686 7.159465 35.31776 0.010624873 0.9733061 8608 -1.645226 6.942654 34.64701 0.010904508 0.9733061 8402 2.224008 6.981127 33.15403 0.011573724 0.9733061 2152 2.173823 6.942653 32.70039 0.011790966 0.9733061 6240 2.568286 6.777594 32.50399 0.011887171 0.9733061 7053 -1.743023 6.777595 32.50192 0.011888190 0.9733061 6984 2.780434 6.942653 32.01302 0.012133576 0.9733061 6057 -1.964729 6.488089 30.77110 0.012797066 0.9733061"
}
]
},
{
"id": "14475",
"date": "01 Jul 2016",
"name": "Nick Schurch",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper lays out a clear and relatively concise example of a linear workflow for analysing an RNA-seq based Differential Gene Expression experiment. The workflow focuses on doing all the analysis steps within R using the author's preferred tools (Rsubread & edgeR) and extends usefully to common aspects of pathway analysis with GO terms, Kegg terms and Gene Set testing.\nWhile I'm sure this paper will be useful for a section of the community (particularly newcomers to this kind of analysis), I find the paper to be quite simplistic and lacking in depth and discussion. In particular, there is no discussion of the subtleties involved in performing this kind of analysis at the coalface of scientific research.\nSome particular points I would like to have seen discussed are:\nHow many replicates should be used for this kind of analysis.\nThe example study uses a very poorly replicated dataset. In this case two replicates per condition *may* be sufficient, but not only is it not discussed but for most experiments this is highly misleading! For new RNA-seq experiments a significantly higher number of replicates should be used, both to guard against problem samples/libraries and to ensure sufficient statistical power to identify significant differential expression (in particular because it is rare to know how large the changes in the data will be before you do the experiment).\n\nHow might one identify problematic issues with datasets that aren't as cooperative as the example dataset.\nFor example, what would significant structure or curvature in the point cloud of the MD plot signify? If the samples don't cluster nicely on the MDS plot by condition, what might this mean (mislabelling, bad replicates, etc)?\n\nHow one might remedy or deal with the problems in such uncooperative datasets.\nFor example, what general approaches could be used for isolating the root cause of the observed problems? How might we adjust the analysis to ameliorate the problems and their downstream impact (dropping datasets, changing mapping parameters, filtering the data)?\n\nHow one might go about choosing sensible selections and thresholds for the data.\nIn my opinion the use of 'standard' and/or 'default' parameters, thresholds and selections (e.g. unique=TRUE, FDR>0.05, log2(FC)>1, cpm>0.5, etc) is a significant and endemic problem in this field. Often these are used solely because they have been used widely before, rather than considering whether they are appropriate for the specific data being analysed. What caused the authors to choose the values they use for this data and what key plots or pieces of information are valuable for choosing these appropriately?\n\nHow the various selection steps, thresholds and even the version of the software used, might impact on the downstream results.\nFor example, if you change the FDR threshold from 0.05 to 0.01, how does this impact the downstream pathway analysis? Are some of these analyses insensitive to threshold values (e.g. gene set analysis) and does this make them better/more useful? If you change the cpm threshold to 1.0, or if you allow non-unique read mappings, how does this impact the number of identified SDE genes and the downstream pathway analyses?\nI am not suggesting that the authors should have given an exhaustive account of the issues. Rather, I think the paper would benefit from briefly discussing some of the more subtle and complex issues surrounding these types of analyses and perhaps highlight some key problems, parameters and thresholds that should be thought about carefully. Without this the paper really presents a very linear, idealized, example of what, in practice, are complex analyses that may require considerably more thought and investigation.\nI also encountered some more specific issues:\nThe link provided does not (currently) link to the actual bioconductor workflow.\n\nI ran all the R commands and they all work (except 'fry' see point 4 below), however I didn't get exactly the same results when (and after) filtering out genes without a symbol.\nThe paper has:\n> head(y$genes)\n\nLength Symbol 497097\n\n3634\n\nXkr4\n\n100503874\n\n3259\n\nGm19938\n\n100038431\n\n1634\n\nGm10568 19888\n\n9747\n\nRp1 20671\n\n3130\n\nSox17 27395\n\n4203\n\nMrpl15\n> y <- y[!is.na(y$genes$Symbol), ] > dim(y) [1] 26357\n\n12\nI had:\n> head(y$genes)\n\nLength Symbol 497097\n\n3634\n\nXkr4\n\n100503874\n\n3259\n\nGm19938\n\n100038431\n\n1634\n\nGm10568 19888\n\n9747\n\nRp1 20671\n\n3130\n\nSox17 27395\n\n4203\n\nMrpl15\n> y <- y[!is.na(y$genes$Symbol), ] > dim(y) [1] 26608\n\n12\nThe change is relatively small but it cascades causing differences in the genes that pass cpm filtering, differences in the normalization factors and differences in the DE results and the downstream analyses. I suspect this is the result of using a slightly older version of org.Mm.eg.db (3.2.3, vs 3.3.0) due to using an older version of R (3.2 vs 3.3). This goes nicely to point (5) above.\n\nThere is no real description of the reasoning behind scaling of the heatmap values to a mean of zero and std dev of one.\n\nThe 'fry' command failed for me, producing the error:\n> fry(y, index=cyt.go.genes, design=design, contrast=B.VvsL) Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), : 'data' must be of a vector type, was 'NULL'",
"responses": [
{
"c_id": "2104",
"date": "02 Aug 2016",
"name": "Gordon Smyth",
"role": "Author Response",
"response": "We thank the reviewer for approving our article. We also thank the reviewer for his comments about RNA-seq analysis in general, but we didn't find the suggestions for revisions to be helpful. One problem is that the reviewer did not follow the instructions given in the article regarding software requirements. In general, Bioconductor workflows and Bioconductor channel articles are intended to be run using the latest release of Bioconductor. This was explicitly explained in the article, which said \"This workflow depends on various packages from version 3.3 of the Bioconductor project, running on R version 3.3.0 or higher.\" The article went on to give the version numbers of all packages used. Unfortunately the reviewer tested the workflow using an earlier version of Bioconductor, with the result that one of the function calls didn't work and there were slight changes in the annotation. We make no apology for providing a \"linear\" workflow demonstrating an easy, robust, fast and flexible workflow from RNA-seq reads through to pathway analysis. That was clearly the purpose of the article. We are pleased that the reviewer found our analysis to be \"idealized\" because it is in fact a typical example of our own biomedical research that we published recently in Nature Cell Biology. The article gives an insight, as far as is possible in a short journal article, of our own analysis process \"at the coalface of scientific research\". The reviewer wants us to take lots of sidepaths, discussing problems that did not in fact occur, but we think that readers will not want to be distracted by hypothetical dead-ends in this way. The instructions for writing these articles (available from https://support.bioconductor.org/p/80077) advised authors to take a pragmatic task-orientated approach and not to get bogged down with extensive discussions of options. One has to start with an example of how an analysis should work in order to have a firm basis from which to deal with problematic studies that might arise in the future. The reviewer also wants us to discuss the consequences of myriad perturbations of thresholds and parameters in the analysis pipeline. On one hand, we are disappointed that the reviewer failed to acknowledge the many explanations that were given in the article. On the other, we think that the specific parameter settings questioned by the reviewer are not particularly crucial and are not the most important issues that we would like researchers to be thinking about when they conduct an analysis. Regarding sample sizes, the minimum appropriate sample size depends very much on the context. n=2 may be sufficient when comparing well sorted cell types from genetically identical mice whereas n=10 may be not nearly enough when comparing whole blood from diseased vs normal patients. For the study analyzed in our article, the cell types have distinct and highly reproducible expression profiles. The results were validated in the biological publication (Fu et al, 2015) in a number of ways. The repeatability of the results was also demonstrated in our current article by the strong correlation between the current results and earlier microarray results on similar cell populations (Figure 9). We disagree with the reviewer's position that sample size recommendations can be made independently of the biological context and the purpose of the scientific study. We now give responses to specific issues raised by the reviewer: 1. How many replicates should be used for this kind of analysis. This is not an article about experimental design. 2. How might one identify problematic issues with datasets that aren't as cooperative as the example dataset. The article already shows users how to create appropriate plots from which problems can be identified. 3. How one might remedy or deal with the problems in such uncooperative datasets. Trying to give a solution to every possible problem that might arise is clearly beyond the scope of the current article. In many cases it might be that no special action needs to be taken as the pipeline we give is quite robust. 4. How one might go about choosing sensible selections and thresholds for the data. What caused the authors to choose the values they use for this data and what key plots or pieces of information are valuable for choosing these appropriately? This has already been explained where relevant in the article. In my opinion the use of 'standard' and/or 'default' parameters, thresholds and selections (e.g. unique=TRUE, FDR>0.05, log2(FC)>1, cpm>0.5, etc) is a significant and endemic problem in this field. Often these are used solely because they have been used widely before, rather than considering whether they are appropriate for the specific data being analysed. The reviewer is entitled to his point of view about the field in general, but all these issues are addressed from first principles in our article. The reason for setting unique=TRUE in the call to align() was explained at the top of page 23. We see no point in unique=FALSE for a gene-level expression analysis. We did not use a 'standard' cpm cutoff but rather explained how to work out a useful threshold for this specific data set. In fact the pipeline is robust to the filtering and tends to give similar results for a range of filtering methods and thresholds, as explained in the article. We provided an extended discussion of DE cutoffs on pages 12-13 (nearly two whole pages). We presented a sophisticated solution using glmTreat() that is better than using a FC cutoff or making the FDR cutoff more stringent. 5. How the various selection steps, thresholds and even the version of the software used, might impact on the downstream results. Already explained where appropriate. For example, page 18 says \"Gene set tests consider all the genes in the specified set and do not depend on any pre-emptive significance cutoff.\" In most cases, changes to the thresholds either have obvious effects (a lower FDR cutoff produces fewer DE genes) or have less impact than the reviewer seems to imply. 1. The link provided does not (currently) link to the actual bioconductor workflow. We submitted our workflow to Bioconductor at about the same time as submitting to F1000Research but it has not yet appeared on the Bioconductor website. Unfortunately, none of the Bioconductor channel articles on F1000Research link to a corresponding code workflow. In the meantime, we have made our code and data available from http://bioinf.wehi.edu.au/edgeR/F1000Research2016 and have added this link to the revised article. Note that the entire LaTeX article is generated automatically by running knit() on a Rnw file. 2. I ran all the R commands and they all work (except 'fry' see point 4 below), however I didn't get exactly the same results when (and after) filtering out genes without a symbol. You used out-of-date versions of R and Bioconductor. The change is relatively small but it cascades causing differences in the genes that pass cpm filtering, differences in the normalization factors and differences in the DE results and the downstream analyses. In fact the DE results are almost identical using either Bioconductor 3.2 or 3.3, demonstrating the robustness of the pipeline. 3. There is no real description of the reasoning behind scaling of the heatmap values to a mean of zero and std dev of one. The rationale was explained (it makes Euclidean distance a function of correlation). We have added one more sentence to the revised article. 4. The 'fry' command failed for me, producing the error: There was no DGEList method for fry in the Bioconductor 3.2."
}
]
},
{
"id": "14478",
"date": "06 Jul 2016",
"name": "Devon P. Ryan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nRNAseq is easily one of the most prevalent NGS experiment types and edgeR one of the most heavily used tools for analyzing the results of these experiments. Given that, I'm quite pleased to see this article from Chen and colleagues that provides a very convenient walk-through of how to perform a typical analysis, including pathway and GO enrichment.\nI have no real reservations regarding this article. Below I'd like to point to a few parts of the paper that could use caveats or further explanation.\nThe example experiment has only two samples per group. That suffices in some circumstances, but at least in my experience the lay reader has the unfortunate habit of reading too much into the the number of samples used in papers like this and then trying to use that as justification for similar sample numbers for their much lower effect size experiments. A caveat or note of warning to those new to RNAseq would have been nice.\n\nThere's typically filtering done, such as the \"rowSums(cpm(y) > 0.5) >= 2\" in this paper. It would have been nice to include some recommendations regarding how to choose a filtering threshold.\n\nThe p-values produced by goana() and topKEGG() are presumably unadjusted for multiple testing. It would have been nice if there had been a note to not then use the typical 0.05 cut off.\n\nWhile playing around with the code presented in the paper, I noticed that the choice of 0.01 for the \"inter.gene.cor\" parameter in camera() has a drastic affect on the resulting p-values. It would be incredibly useful to know when one should override the default value and how one should then derive an appropriate value. My concern is primarily that many will see these commands as \"the one true method\" for performing such an analysis and blindly apply the option in cases where it might not be appropriate.",
"responses": [
{
"c_id": "2101",
"date": "02 Aug 2016",
"name": "Gordon Smyth",
"role": "Author Response",
"response": "Thank you for your thoughtful report. We have added a note about small sample numbers to the end of the section describing the experiment. To be honest, we thought that we had already explained how to choose the filtering threshold in some detail. Admittedly we explained in words how the 0.5 and the 2 values in the filter formula were derived, and why they are appropriate for this experiment, rather than giving a formula. Anyway, we have edited the filtering section and added a couple of sentences. We have added a note about the p-values from topGO() and topKEGG(). We generally ignore p-values above about 1e-5. You ask a good question about inter.gene.cor for camera(). We have recently made inter.gene.cor=0.01 the default setting for camera(). Previously the default was to estimate the correlation separately for each gene set. The old default gives rigorous control of the type I error rate but is conservative and doesn't always rank the most biological interpretable sets most highly. The ranking issues occurs because of the need to penalize highly co-regulated sets with positive inter-gene correlations. We and others (Tarca et al, 2013) have noticed that much simpler methods like limma's geneSetTest() tend to give a better ranking of the biologically significant sets although they do not control the error rate correctly. Our recent use of a preset value for inter.gene.cor in camera() is an attempt to strike a compromise between the original camera() and geneSetTest(). Note the latter is equivalent to camera() with inter.gene.cor=0. The compromise gains the advantages of geneSetTest() while keeping reasonable, although not perfect, error rate control. You are right that the camera p-values are sensitive to the value for inter.gene.cor. Nevertheless, after quite a bit of experimentation, we have chosen the value of 0.01 as a reasonable compromise between ranking and error rate control that gives good results across a range of datasets. So we are happy to offer it for general use and would prefer that most users kept to the default value. The p-values will often be somewhat optimistic, but probably not more so than other commonly used methods like the Fisher tests for GO and KEGG terms. It gives vastly better error control than gene sets methods that permute genes and ignore inter-gene correlations, especially for larger sets. Reference Tarca AL, Bhatti G, Romero R (2013) A comparison of gene set analysis methods in terms of sensitivity, prioritization and specificity. PLoS ONE 8(11): e79217. http://dx.doi.org/10.1371/journal.pone.0079217"
}
]
},
{
"id": "14480",
"date": "13 Jul 2016",
"name": "Tsung Fei Khang",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFirst of all, please accept my apologies for not being able to complete the review earlier. I would like to thank the editor for the opportunity to review this interesting paper. EdgeR is one of the more popular methods for performing RNA-Seq data analysis, and the authors’ efforts in writing this expository tutorial will surely be welcome by students and researchers who need to work with analysis of RNA-Seq data.\nI think entry-level readers with basic R skills will find the workflow described easy to follow; and having actually worked out an example data set, will have less difficulty in adapting it to the needs of their own data analysis. However, intermediate or experienced readers might find the presentation of some parts of the workflow overly simplistic (as pointed out by one of the reviewers N.J.Schurch) - but this is not a weakness of the paper, since it is too much to expect that all the nuances of a refined RNA-Seq data analysis can be covered in this tutorial paper. Nonetheless, I believe the discussion of these points will potentially add value to the paper – they do not necessarily imply the necessity of revising the work to include the points raised, because the appending of reviewers’ comments enables readers to assess how relevant are the points raised to their own work.\nTo ease discussion of the paper, I will itemize my comments as follows.\nDespite declaring “From reads to genes to pathways” in their title, the authors choose not to develop the contents of their paper in this sequence. Rather, they start immediately with the count table, and then develop the material going from genes to pathways. The section on read mapping is presented at the end instead.\nI understand the focus of their sequence of writing, which is to get the readers into the heart of the action quickly, using the tool that they are most familiar with, edgeR. However, read processing is an important upstream checkpoint, and the things that one chooses to do at this stage has more important consequences than choices of which contrasts to make, optimizing plots, etc. Personally, I feel that insufficient attention has been given to this section, which can benefit from more discussion. It troubles me that there is no mention of short read quality control, a standard (and important) requirement for data quality check, which I am sure the authors are aware of. This is typically visualized using the standard tool FastQC. Subsequently, depending on the diagnostics, one can use a tool like FastX or Trimmomatic1 to remove problematic segments, usually the 3' end. Then, there is a plethora of read mapping methods that can be used (of which Rsubread is just one of them), and also methods of constructing the count table from the mapped reads. Optimal combinations of methods for performing both tasks were recently investigated by Fonseca et al.2, who suggested combinations such as OSA+HTSeq for producing the reliable count tables. While the authors do not really need to show how these can be done, I think they should devote a short paragraph to discuss these issues because of their fundamental nature.\n\nThe authors demonstrate the use of mean difference (MD / a.k.a MA) plots as a diagnostic plot for checking data distributional properties. These are useful for checking whether variances increase as counts get larger in samples (the “fanning patterns”), for example. Less clear is the appropriate course of action in the event observing such undesirable patterns. Do we try to carry on the analysis, using log-transformed data? Do we discard problematic samples? Admittedly these are delicate issues that require more space for discussion than is possible in the paper. Nevertheless, providing some guidelines or pointing out useful references for further exploration will surely help readers appreciate the use of these plots.\n\nThere is strangely no illustration of how to make a volcano plot in the tutorial, which is a common graphical plot for assessing the joint relationship between statistical and biological significance. From experience, I find such plots important for understanding how different DE methods pick DE gene candidates.\nI was motivated to understand how glmQLFTest and glmTreat functions call DE genes compared to a simple method based on hyperbolic decision rules proposed by Xiao et al.3 using the volcano plot. The method of Xiao constructs the decision rule as follows: Declare a fold change (FC) cut-off below which one is not interested in a gene as a DE candidate. So if we desire FC > 2 for up-regulated genes (and conversely FC < 1/2 for down-regulated genes), then |log2(FC)| > 1. Next, we set the level of statistical significance, above which a gene is considered to be an unlikely DE candidate. Suppose we use the adjusted p-value, and require p < 0.01. This implies that –log10(p) > 2. If we denote y = -log10(p) and x = log2(FC), then the product of these two inequalities gives |x|y > 2, so that y > 2/|x|. This translates to a hyperbolic decision rule, such that genes with x and y values lying in the rejection region are selected as DE candidates. This rule allows one to include genes with very large FC but higher p-value. If we care a lot about managing false positives, then we could add a hard requirement for –log10(p) = 2, meaning that we will only considering genes that demonstrate adjusted p-values below 0.01.\nThe result of my exploration is attached here. Non-DE genes are in black. The genes picked using glmQLFTest are in green; note the majority of them are also picked by the hyperbolic decision rule (blue). Candidates returned from glmTreat are boxed in purple, and seems to form a subset of the candidates returned from the hyperbolic decision rule. However, they show a peculiar distribution pattern, in that some genes with large log2(FC) and –log10(p) do not get picked. It is unclear to me why such genes are not detected by the algorithm. Regardless, a volcano plot is an important instrument that readers can have at their disposal for understanding the behaviour of DE gene calling algorithms.\n\nThe heatmap (Fig.7) is an important graphical plot of any gene expression analysis project, but there are some subtleties to its proper generation. I think it is not easy to explain the clustering pattern of the samples in Fig. 7, where basal and luminal cell samples are grouped together. Fortunately, this is often just a problem of the choice of clustering algorithm used. The default method in heatmap.2 for clustering is complete linkage, which is often not the best method. From experience, changing it to the ward.D algorithm frequently produced biologically meaningful results, which is the case in the current analysis.\nThe figure here shows a possible modification of the heatmap. Here, the basal and luminal samples nicely separate out into two clusters, following biological intuition. Columns of interest (e.g. lactating state in both basal and luminal cells) can be boxed to draw reader’s attention.\nMay I also recommend that the srtCol argument in the heatmap be introduced to users, since sooner or later one would have to deal with space issues with labels on a heatmap, and what better way to handle this than having them oblique instead of perpendicular to the plot? Additionally, I think the outcome of customizing the heatmap using the given margin, lhei and lwid arguments will produce variable results in different computers (I got a \"figure margins too large\" error message initially), and so a note to users may be useful.\n\nIn the output table produced using the topTags function, there is a column named “FDR”. Should this be “Adjusted p-value”? In the Benjamini-Hochberg correction method that the authors’ used, FDR is a parameter determined by the user. Depending on one’s taste for false positive tolerance, one can tune it low or high (maybe useful to let users tune it?), so synonymizing “FDR” with “adjusted p-value” leads to conceptual confusion. By the way, how did the authors compute the adjusted p-value?\n\nWhile testing out the codes, I noted that the output that I got differed slightly from those shown by the authors. Additionally, like N.J.Schurch, I also encountered problems running the fry code example:\n> fry(y, index=cyt.go.genes, design=design, contrast=B.VvsL) Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), : 'data' must be of a vector type, was 'NULL'\nOnly much later in the end did I read, on page 21, that the authors made their analysis using R version 3.3.0 or higher, with Bioconductor version 3.3. Since my versions for both were 3.2, I suppose that the variation in output, as well as the error message seen, could be just a consequence of different versions. Would it be better if the versions used are announced right at the beginning of the paper? Additionally, it would also help if the packages needed for running the analysis are all installed at the beginning of the R script provided (e.g. readers who had not run biocLite(“GO.db”) to install the package from Bioconductor would get an error running library(GO.db) – this is not mentioned in the text I think, and can trouble beginners)\n\nMinor comments:\n(i)Subject-verb agreement issue (page 6): “We require that gene(s) have a count ….”.(ii)ANOVDEV - analysis of deviance, a citation is useful.\n\nNote: Codes for producing the volcano plot and heatmap are available here.",
"responses": [
{
"c_id": "2103",
"date": "02 Aug 2016",
"name": "Gordon Smyth",
"role": "Author Response",
"response": "We thank the reviewer for his thoughtful comments, although we have different views on some of the specific issues raised. The reviewer is right that this is a tutorial aimed at entry-level readers. We went to some trouble to make the tutorial simple, but we do not think it is \"simplistic\". The article alludes to many data analysis issues in a concise style. It includes material that is new and interesting for an advanced reader, as the reviewer reports show. Our article is \"live\" in the sense that readers can regenerate the analysis and the article themselves automatically from a knitr file containing R code. In this revision, we have added a link to the code files. It isn't practical to include the read alignment in a live analysis because it is so much more computationally demanding than the rest of the analysis. Requiring readers to undertake the alignment step to obtain the read counts would drastically limit the audience who could go through the rest of the analysis. We have added a note to the article about plotting quality scores. However we think that trimming poor quality segments is an old-fashioned step that is generally unnecessary given improved sequencing protocols and high quality robust aligners like subread. Trimming is more likely to be harmful than helpful for a gene-level RNA-seq study. Indeed we think that encouraging entry-level users to make ad hoc edits of their sequence data is quite dangerous. It is far better to allow a quality-score-aware aligner like subread to make decisions on a per-read basis. We are not sure why the reviewer cites Fonseca et al (FMB), but we make the following points. FMB evaluated pipelines for quantifying absolute expression, which is not directly relevant in a differential expression study such as ours. FMB only compared pipelines available at the time. The OSA+HTSeq pipeline is not particularly popular, for example it has not been adopted by any of the FMB authors themselves for any published study of real data. By contrast, the Rsubread+featureCounts pipeline that we suggest is newer, faster and more widely used. We could cite references to claim superiority for the Rsubread and featureCounts tools, for example the SEQC study (Su et al, 2014), but a review of the literature would be out of place in our article. What is undoubtedly true is that Rsubread+featureCounts is more than good enough and easily the fastest and most convenient in an R context because of its native implementation as an R package. The reviewer may have misinterpreted the purpose of the MD plots. They are designed to display differential expression, either for individual samples or for a fitted model. They are not designed to check distributional properties. They do not check whether variances increase with count size. They are not used to suggest transformations of the data. Volcano plots were originally motivated by the shortcomings of ordinary t-tests, which can give very small p-values even for genes with tiny fold changes. However this problem has been overcome by empirical Bayes test statistics, and we do not generally recommend volcano plots in the context of an edgeR analysis. Volcano plots tend to encourage fold change cutoffs, which we also don't recommend. We much prefer the MD plot (Figure 5) because it shows clearly how larger fold changes are required to reliably call DE for lower expressed genes. The decision rule of Xiao et al (2014) doesn't give rigorous control of the FDR, and it has the tendency to prioritize genes with small counts that have large fold changes and large variances. We prefer not to prioritize lowly expressed genes. Again, this is made clear by the MD plot. Thank you for the suggestions and code for the heatmap. The pattern seen in our heatmap is because we chose to display genes that are DE between B.pregnant and B.lactating, hence it is natural that these two populations are separated at the far left and right of the plot. We agree that slanted labels are useful, as are some of the other options you demonstrate. The choice of clustering algorithm is controversial, especially so as the ward.D algorithm is not a correct representation of Ward's method, with some writers claiming that only ward.D2 should be used. Anyway, our article is not about heatmaps per se. Our aim is simply to provide an example of how results can be transferred to a heatmap, so it is best to keep the heatmap call as simple as possible. Users are then free to add as many embellishments as they like. There is no conceptual confusion with FDRs. For any FDR tolerance that a user might have, the genes for which the FDR value from topTags() is less than this cutoff are exactly the same genes that would be judged statistically significant by Benjamini-Hochberg's 1995 algorithm. Obviously the results from the workflow will differ slightly if older versions of R and Bioconductor are used. The software versions were stated on page 18 as well as on page 21. The journal format is that software requirements are described at the end of the article. The requirements seem to us to be well sign-posted in sections called \"Packages used\" and \"Data and software availability\". In any case, we are a bit surprised that readers should need special prompting to install the current versions of R and Bioconductor. Package installation is a \"once off\" operation, so we prefer not to make it part of the workflow code that a user might run many times. A citation for ANODEV has been added. Reference Su, Z, et al (2014). A comprehensive assessment of RNA-seq accuracy, reproducibility and information content by the Sequencing Quality Control Consortium. Nature Biotechnology 32(9), 903-914."
}
]
},
{
"id": "14474",
"date": "18 Jul 2016",
"name": "Steve Lianoglou",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral Comments\nI'd like to first thank the authors for having a long history of providing key contributions to the filed of bioinformatics in terms of methodological advances which are manifested in robust software (limma and edgeR), and further for being most generous with their time by providing extensive support for their software and its use by writing (epic) user guides and answering an innumerably many questions on the bioconductor support forum -- especially since the latter is likely not considered \"important\" (citable(!)) under most models of academic recognition. The community owes you a large debt of gratitude.\nNow, for this workflow: The authors have provided a complete tutorial on the analysis of RNA-seq data that addresses many of the considerations required while performing these tasks. I'm particularly happy to see that the authors draw people's attention to the use of their \"treat\" framework, the brief (but important) ANODEV section, as well as discussing different ways to perform gene set enrichment analysis.\nMy strongest comment is that this is very well written and should prove very useful to the community at large -- and most useful to the \"casual\" analyst, one who isn't well versed in the various avenues of research that are now so conveniently wrapped up behind a single call to `glmTreat` or `camera`. In this vein, I feel the authors have done a commendable job of touching upon many of the more subtle parts of the data preparation steps in an RNA-seq analysis (ie. the importance of filtering, how normalization factors are used to adjust library size, explanation of an MDS plot, etc).\nOf course one can't comment on every corner case that might arise during an rna-seq analysis, but to benefit this audience most, I'd like to point out some points that I feel could benefit from further clarification. Other readers would likely wish the authors clarify another set of points. The set of points that are most important to discuss is going to be subjective and based on our own experience in analyzing datasets (and helping others do the same), but for me I feel at least these could use some more elaboration:\nAfter the `plotMD` code example, the authors mention that \"the bulk of genes should be centered at a line of zero log-fold change ...\", it might be worth mentioning a few options to consider when a vanilla call to `calcNormFactors` doesn't produce that result.\n\nIt is ultimately the user's responsibility to keep up with the primary publications in the field, but I think the authors can help with just a few clarifying comments.\n\nIn the Introduction, the authors cite [12] (A. Lun, et al. It's DE-licious ...) when they mention edgeR's QLF framework offers some additional statistical refinements when compared to QuasiSeq, but [12] doesn't seem to mention any direct comparisons to QuasiSeq at all. As far as I can tell, [12] only mentions that QuasiSeq also uses quasi-likelihood F-tests, and that these account for gene-specific variability from both biological and technical sources. Could the authors clarify what these refinements might be? (In retrospect, this comment seems a bit trivial to make. I intended to chase more comments to publications, but unfortunately don't have the time ... I doubt that there's any need to, just trying to help the casual analyst connect some dots here, is all)\n\nFrom previous experience of putting camera's `inter.gene.cor` parameter to use, I can say it's both awesome and mysterious. Awesome because camera's rankings in this mode are often very useful, but mysterious because: what do the p-values now mean, really? How much should the analyst care? The original camera paper goes to some length to discuss the importance of type I error control and that camera's approach of estimating and accounting for inter-gene correlation is an improvement there. Given that the user can now override it, what can the analyst reasonably say about type I error control, now? Some guidance on choice of the value, or (perhaps) comment on why it's not so critical could be useful.\n\nMinor Comments\nThe authors have done a good job of enabling easy reproducibility of this workflow. Keeping with that spirit, it might be useful to change the code that materializes the `targets` object to be executable without the use of an external file. Leveraging R's ability to read from a `textConnection` might be one day to do that without loosing readability of the workflow, since the targes file would also be printed to the document without having to output its value:\n\ntargets <- read.delim(textConnection(\"\n\nGEO\n\nSRA CellType\n\nStatus\n\nMCL1.DG GSM1480297 SRR1552450\n\nB\n\nvirgin\n\nMCL1.DH GSM1480298 SRR1552451\n\nB\n\nvirgin\n\nMCL1.DI GSM1480299 SRR1552452\n\nB pregnant\n\nMCL1.DJ GSM1480300 SRR1552453\n\nB pregnant\n\nMCL1.DK GSM1480301 SRR1552454\n\nB lactating\n\nMCL1.DL GSM1480302 SRR1552455\n\nB lactating\n\nMCL1.LA GSM1480291 SRR1552444\n\nL\n\nvirgin\n\nMCL1.LB GSM1480292 SRR1552445\n\nL\n\nvirgin\n\nMCL1.LC GSM1480293 SRR1552446\n\nL pregnant\n\nMCL1.LD GSM1480294 SRR1552447\n\nL pregnant\n\nMCL1.LE GSM1480295 SRR1552448\n\nL lactating\n\nMCL1.LF GSM1480296 SRR1552449\n\nL lactating\n\n\"), sep=\"\")\n\nIn the \"Downloading the read counts\" section, the authors say that the first column of the downloaded read counts is the total number of bases in exons or UTRs for the gene, but UTRs *are* exons (sorry, pet peeve of mine) -- perhaps \"total number of of basepairs from coding and non-coding exons(?)\"\n\nWhen construction the DGEList, why not show that you can now easily drop the `targets` data.frame into the DGEList's `$samples` slot like so:\n\ny <- DGEList(GenewiseCounts[,-1], group=group, samples=targets,\n\ngenes=GenewiseCounts[,1,drop=FALSE])\n\nGrammar / Spelling\n\nIn the \"Filtering to remove low counts\" section\n\n- 'Genes that have *with* very low counts across all the libraries ...'\n\n- As a rule of thumb, we require that gene have a count of ...'\n\n+ fix this sentence in a few places to support the use of \"gene\" or \"genes\"\n\nAfter `plotBCV`: \"The dispersion trend tends [to] decrease smoothly with abundance and (to -> is) asymptotic to a constant ...\"\n\nWhen the concept of dispersion estimation is introduced, the different types of dispersions are briefly discussed (each individual gene, common dispersion, trended dispersion of a gene). Then, in the QL approach, you mention that the \"tagwise NB dispersions\" are not used. Would be useful to use same naming convention (ie. which of the previous types of dispersions introduced is the tagwise dispersion referenced here?)\n\nRemove third \"the\" here: \"The QL functions moderte the genewise the QL dispersion estimates ...\"\n\nUnder *complicated contrasts*: Suppose we are interested in testing whether the change in expression between lactating and pregnant mice is the same for basal cells [as] it is for luminal cells.",
"responses": [
{
"c_id": "2102",
"date": "02 Aug 2016",
"name": "Gordon Smyth",
"role": "Author Response",
"response": "Thanks, Steve, for your positive comments. Here are some responses to your extra suggestions. After the `plotMD` code example, the authors mention that \"the bulk of genes should be centered at a line of zero log-fold change ...\", it might be worth mentioning a few options to consider when a vanilla call to `calcNormFactors` doesn't produce that result. A paragraph has been added at the end of the normalization section. The MD plots have been moved to the next section with some added discussion of the relationship to normalization factors. In the Introduction, the authors cite [12] when they mention edgeR's QLF framework offers some additional statistical refinements when compared to QuasiSeq ... Could the authors clarify what these refinements might be? Robust empirical Bayes (robust=TRUE) is one refinement. Another is more careful treatment of zero counts when modelling the residual deviances. See our response to reviewer 1 (Conrad Burden). From previous experience of putting camera's `inter.gene.cor` parameter to use, I can say it's both awesome and mysterious. ... Some guidance on choice of the value, or (perhaps) comment on why it's not so critical could be useful. See our response to reviewer 2 (Devon Ryan). The authors have done a good job of enabling easy reproducibility of this workflow. Keeping with that spirit, it might be useful to change the code that materializes the `targets` object to be executable without the use of an external file. Leveraging R's ability to read from a `textConnection` might be one day to do that without loosing readability of the workflow That's a interesting suggestion. Alternatively of course we could have simply created the CellType and Status vectors in R without reading a targets frame at all. The reason why we read from an external file is we feel that doing so has pedagogic value. We think it is generally good practice for experimenters to create the targets file outside of R, using a spreadsheet editor like Excel. This forces the experimenter to check the correspondence between sample IDs and experimental factors. We want the workflow to mimic how a real analysis will go. In the \"Downloading the read counts\" section, the authors say that the first column of the downloaded read counts is the total number of bases in exons or UTRs for the gene, but UTRs *are* exons We agree, but the text nevertheless seems simple and clear. In our experience, it helps to explicitly mention UTRs. We've changed it to \"exons and UTRs\". When constructing the DGEList, why not show that you can now easily drop the `targets` data.frame into the DGEList's `$samples` slot like so: Good suggestion. We don't do it just so that the output from y$samples on pages 4 and 5 doesn't exceed the window width. Grammar / Spelling Fixed. When the concept of dispersion estimation is introduced, the different types of dispersions are briefly discussed (each individual gene, common dispersion, trended dispersion of a gene). Then, in the QL approach, you mention that the \"tagwise NB dispersions\" are not used. Would be useful to use same naming convention (ie. which of the previous types of dispersions introduced is the tagwise dispersion referenced here?) Good suggestion, done."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1438
|
https://f1000research.com/articles/5-1898/v1
|
01 Aug 16
|
{
"type": "Research Article",
"title": "The use of collagen matrix (Ologen) as a patch graft in glaucoma tube shunt surgery, a retrospective chart review",
"authors": [
"John D. Stephens",
"Steven R. Sarkisian, Jr.",
"Steven R. Sarkisian, Jr."
],
"abstract": "Purpose: To determine the safety and efficacy of collagen matrix as a patch graft in glaucoma drainage surgery. Collagen matrix grafts may be advantageous because they do not need to be harvested from human donors. Methods: An institutional, retrospective review of 43 patients with at least 12 months follow-up status post-glaucoma drainage implant surgery were evaluated for signs of tube erosion after initial placement of collagen matrix patch graft. Results: Forty-one of 43 eyes (95.3%) required no intervention for patch graft melting with tube erosion. Average time of follow-up was 32 months (range: 12-45). Two cases had tube erosion at 4 months and 26 months post-op requiring tube revision, which was successfully revised with conjunctiva (4 month erosion) and donor sclera (26 month erosion). Conclusion: Our results suggest that collagen matrix patch grafts may be used successfully as a patch graft in glaucoma tube shunt surgery, and may be advantageous because they do not have to be harvested from human donors. It is possible that exposure rates may be higher after longer follow-up and with larger numbers of patients. Further research is needed to compare Ologen to traditional graft materials to conclusively determine the safety and efficacy of collagen matrix as a novel patch graft material.",
"keywords": [
"Ophthalmology",
"glaucoma",
"glaucoma surgery",
"glaucoma tube shunt",
"glaucoma patch graft",
"collagen matrix"
],
"content": "Introduction\n\nThe use of glaucoma drainage implants to treat difficult glaucoma cases has increased in the past two decades1. These devices drain aqueous through a silicone tube to a reservoir plate covered by Tenon’s capsule and conjunctiva. The tube is then covered by one of several materials to prevent exposure to the overlying conjunctiva. Although most complications are transient and self-limited, glaucoma drainage procedures carry the risk of persistent corneal edema, tube erosion, endophthalmitis/blebitis, and tube migration, among other complications2. Tube shunts in particular carry the risk of patch graft thinning and exposure of the subconjunctival portion of the shunt tube, which is a risk factor for infectious endophthalmitis3,4. Prompt identification and revision of exposed patch grafts with collagenous human autograft or allograft material is therefore recommended5.\n\nSeveral patch graft materials have been used. These include pericardium, fascia lata, cornea, sclera, and amniotic membrane6,7. Ologen (Aeon Astron Europe BV, Leiden, the Netherlands) is a porcine-derived biodegradable collagen matrix implant which has been studied and used as an adjunct to trabeculectomy8,9. A recent case report showed successful use of Ologen as a patch before closing the conjunctiva in a case of tube erosion10. To our knowledge, Ologen has not been used as a primary patch graft in glaucoma tube shunt procedures. Collagen matrix may be advantageous because it does not need to be harvested from human donors and is less expensive than other patch graft materials. This is particularly important considering that Medicare (the federal health insurance program for people who are 65 or older, medicare.gov) now no longer reimburses for any patch graft material when combined with a tube shunt procedure (former CPT code 67255). Additionally, Ologen appears clear under the conjunctiva and provides improved cosmesis compared to other patch grafts (Figure 1, printed with permission courtesy of Steven R. Sarkisian, jr.). The purpose of this study was to determine the safety and efficacy of collagen matrix as a patch graft in glaucoma tube shunt surgery.\n\nBlack arrow: tube in anterior chamber. Blue arrow: Ologen patch graft.\n\n\nMaterials and methods\n\nThis study was approved and monitored by the Institutional Review Board at the University of Oklahoma Health Science Center (IRB# 3425; reference #652312). Permission to publish clinical details and images was obtained for each subject. Potential subjects were identified by reviewing case logs of a single attending surgeon (S.R.S.). Charts of consecutive patients undergoing glaucoma tube shunt surgery with placement of collagen matrix patch graft between July 2009 and December 2010 were reviewed. Charts were excluded if the patient had less than 12 months of follow-up data. Forty-three eyes of 40 patients were identified. Demographic and clinical information of the patients is listed in Table 1. The primary outcome measure of this study was post-operative tube exposure requiring revision.\n\n\nSurgical technique\n\nThe glaucoma drainage implant of choice was placed in the usual fashion11. Once the tube was secured to the sclera, the collagen patch graft was used to cover the tube (Figure 2). The Ologen to cover a tube comes as a 10×10×2 mm sheet. Presoaking the collagen is not necessary and is, in fact discouraged because once wet, the collagen becomes difficult to cut and can tear easily. While dry, the collagen sheet was cut to size to cover the tube per the surgeon's preference. Although some surgeons may desire to suture the collagen in place, we find this unnecessary as the collagen quickly picks up moisture from the scleral bed, does not slide out of place easily and never moves post-operatively once the conjunctiva is closed. However, great care is taken to ensure that the collagen is fully covered and the conjunctiva covering it is not under tension. Every effort must be made to be certain there is no chance that any part of the collagen is exposed and the conjunctiva is well secured. Once the conjunctiva was closed, a small amount of saline was placed in the anterior chamber and a fluorescein strip was used to verify the absence of leakage.\n\nRed arrow: tube in anterior chamber. Blue arrow: conjunctiva over patch graft. Orange arrow: Ologen patch graft.\n\n\nResults\n\nA brief summary of results is displayed in table 2. Forty-one of 43 (95.4%) eyes with Ologen patch graft required no intervention for patch graft melting with tube erosion. The average time of follow-up was 32 months (range 12–45 months). Two cases had tube erosion requiring revision. These occurred at 4 months and 26 months post-operatively. The first patient was an 86-year-old Caucasian woman with open angle glaucoma and a history of iritis. She had partial exposure of the patch graft after 1 week and full exposure at 4 months. She underwent successful tube revision with conjunctiva for a total follow up of 32 months. The second erosion occurred in a 74-year-old Caucasian woman with open angle glaucoma and long-standing diabetes mellitus. The erosion occurred at 26 months and was successfully repaired with donor sclera for a total follow up of 32 months. Neither patient developed signs of endophthalmitis during their clinical course. Both of these patients had Ahmed valves placed in the superotemporal quadrant. One patient in this study, a 63-year-old man with open angle glaucoma, developed partial tube exposure on post-operative day 10 but did not require revision. He underwent placement of Baerveldt shunt in the inferonasal quadrant.\n\n\nDiscussion\n\nTo our knowledge, no study has investigated the use of collagen matrix material as a primary patch graft in glaucoma tube shunt surgery. Previous studies have reported rates of patch graft erosion. Gedde et al. reported tube erosion in five of 107 eyes (4.6%) in the tube versus trabeculectomy study at 5 years of follow-up11. In a study of 702 patients, Levinson et al. reported an exposure rate of 5.8% at a mean follow up of 36 months12. Additionally, Muir et al. reported an exposure rate of 6.2% in 1073 patients followed for an average of 41 months13. The erosion rate in our study, 4.7%, is comparable to these previous studies.\n\nSeveral factors may predispose patients to patch graft erosion. In a cohort study of 121 eyes, Koval et al. identified Hispanic ethnicity, neovascular glaucoma, previous trabeculectomy, and combined surgery as potential risk factors for tube shunt exposure14. In the aforementioned study by Muir et al., female gender and white race were associated with an increased risk of graft exposure. Uveitis, diabetes, and type of tube shunt were not associated with increased risk13. Mechanical and immunologic factors may also contribute to graft erosion15. Both of the patients with graft erosion in our study had histories suggestive of poor wound healing and/or ocular inflammation. One had long-standing diabetes mellitus without a diagnosis of neovascular glaucoma. The second patient with erosion in our study had a history of iritis.\n\nOlogen encapsulates when not exposed to aqueous and does not biodegrade. It is possible that the patch graft erosions in our study occurred because the Ologen was exposed and not well-covered initially, leading to patch melting. Care must be taken to not use Ologen if the conjunctiva is under tension when it is closed.\n\nThere are several limitations to this study. First, given its relatively small sample size and limited duration, further studies are necessary to determine the safety and efficacy of Ologen collagen matrix patch grafts compared to other commonly used materials. There are inherent limitations in a retrospective chart review, including lack of randomization of patients, lack of comparative control group and incomplete follow-up by patients not reviewed for this study. A prospective, large, controlled study is needed to compare erosion rates of Ologen to other graft materials. It is possible that collagen matrix patch grafts may be used successfully in glaucoma tube shunt surgery. They may be advantageous because they do not need to be harvested from human donors, are less expensive, and provide improved cosmesis compared to other commonly used materials. Further study is required to evaluate the long-term use of Ologen as a patch graft.\n\n\nData availability\n\nF1000Research: Dataset 1. The use of collagen matrix (Ologen) as a patch graft in glaucoma tube shunt surgery, a retrospective chart review data spreadsheet. 10.5256/f1000research.9232.d13089416\n\n\nEthical considerations\n\nThis study was approved and monitored by the Institutional Review Board at the University of Oklahoma Health Science Center (IRB# 3425; reference #652312). Permission to publish clinical details and images was obtained for each subject.",
"appendix": "Author contributions\n\n\n\nSRS conceived the study and performed the procedures reviewed in this study. JS carried out the research, analyzed the results, and prepared the first draft of the manuscript. JDS and SRS were involved in the revision of the draft and manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nDr. Sarkisian is a consultant for Alcon, Aeon Astron, Beaver Vistec, InnFocus, Sight Sciences, New World Medical, and Ellex He is currently receiving research grants from Alcon, Aeon Astron, Aerie, Transcend, Glaukos, and is a lecturer for Alcon.\n\n\nGrant information\n\nThis paper was funded in part by Research to Prevent Blindness.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nGedde SJ, Schiffman JC, Feuer WJ, et al.: Treatment outcomes in the tube versus trabeculectomy study after one year of follow-up. Am J Ophthalmol. 2007; 143(1): 9–22. PubMed Abstract | Publisher Full Text\n\nGedde SJ, Scott IU, Tabandeh H, et al.: Late endophthalmitis associated with glaucoma drainage implants. Ophthalmology. 2001; 108(7): 1323–1327. PubMed Abstract | Publisher Full Text\n\nGedde SJ, Herndon LW, Brandt JD, et al.: Surgical complications in the Tube Versus Trabeculectomy Study during the first year of follow-up. Am J Ophthalmol. 2007; 143(1): 23–31. PubMed Abstract | Publisher Full Text\n\nKrebs DB, Liebman JM, Ritch R, et al.: Late infectious endophthalmitis from exposed glaucoma setons. Arch Ophthalmol. 1992; 110(2): 174–175. PubMed Abstract | Publisher Full Text\n\nKalenak JW: Revision for exposed anterior segment tubes. J Glaucoma. 2010; 19(1): 5–10. PubMed Abstract | Publisher Full Text\n\nSmith MF, Doyle JW, Ticrney JW Jr: A comparison of glaucoma drainage implant tube coverage. J Glaucoma. 2002; 11(2): 143–147. PubMed Abstract | Publisher Full Text\n\nAnand A, Sheha H, Teng CC, et al.: Use of amniotic membrane graft in glaucoma shunt surgery. Ophthalmic Surg Lasers Imaging. 2011; 42(3): 184–189. PubMed Abstract | Publisher Full Text\n\nJohnson MS, Sarkisian SR Jr: Using a collagen matrix implant (Ologen) versus mitomycin-C as a wound healing modulator in trabeculectomy with the Ex-PRESS mini glaucoma device: a 12-month retrospective review. J Glaucoma. 2014; 23(9): 649–52. PubMed Abstract | Publisher Full Text\n\nDada T, Kusumesh R, Bali SJ, et al.: Trabeculectomy with combined use of subconjunctival collagen implant and low-dose mitomycin C. J Glaucoma. 2013; 22(8): 659–62. PubMed Abstract | Publisher Full Text\n\nRosentreter A, Schild AM, Dinslage S, et al.: Biodegradable implant for tissue repair after glaucoma drainage device surgery. J Glaucoma. 2012; 21(2): 76–78. PubMed Abstract\n\nChristakis PG, Tsai JC, Zurakowski D, et al.: The Ahmed Versus Baerveldt study: design, baseline patient characteristics, and intraoperative complications. Ophthalmology. 2011; 118(11): 2172–9. PubMed Abstract | Publisher Full Text\n\nGedde SJ, Herndon LW, Brandt JD, et al.: Postoperative complications in the Tube Versus Trabeculectomy (TVT) study during five years of follow-up. Am J Ophthalmol. 2012; 153(5): 804–814.e1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevinson JD, Giangiacomo AL, Beck AD, et al.: Glaucoma drainage devices: risk of exposure and infection. Am J Ophthalmol. 2015; 160(3): 516–521.e2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuir KW, Lim A, Stinnett S, et al.: Risk factors for exposure of glaucoma drainage devices: a retrospective observational study. BMJ Open. 2014; 4(5): e004560. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoval MS, El Sayyad FF, Bell NP, et al.: Risk factors for tube shunt exposure: a matched case-control study. J Ophthalmol. 2013; 2013:1–5, 196215. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStephens J, Sarkisian SR Jr: Dataset 1 in: The Use of Collagen Matrix (Ologen) as a Patch Graft in Glaucoma Tube Shunt Surgery, a Retrospective Chart Review. F1000Research. 2016. Data Source"
}
|
[
{
"id": "15375",
"date": "08 Aug 2016",
"name": "Nils Loewen",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOlogen is an FDA approved biodegradable collagen matrix that can be used in trabeculectomy to maintain the bleb space and modulate wound healing1,2 and repair of scleral defects3. No lyophilized or prepared donor tissue is required which potentially has cost, safety, storage and standardization advantages. The authors are likely some of the most experienced surgeons worldwide with this material and need to be commended for sharing their experience and insight. In this retrospective study, Stephens et al evaluated the outcomes of this porcine derived biodegradable collagen matrix as a patch graft in tube shunt surgery. The authors found that 2 eyes out of 43 had tube erosion and required further surgical interventions. Both patients had a history of poor wound healing or ocular inflammation increasing the risk of erosion. It is intriguing that the rate of erosion is lower, 4.7%, compared to other materials or at least similar (indicating non-inferiority). This first description of Ologen as a tube patch material is interesting and useful and might standardize tube shunt surgeries while avoiding human donor tissue.\nI believe this already good manuscript can be made more useful by adding the following:\nCan the authors please report the IOP data before and after surgery as well as number of medications?\n\nPower calculation: approximately how many patients would be needed to show that ologen is better than the reported rates for other materials? While not critical it is possible that the patient number is large enough to make this statement and if not, at least non-inferiority is likely. In my opinion, is not a critical issue that should not prevent this article from being approved.\n\nOlogen handling: the authors are extremely experienced with how ologen behaves and a description in the discussion would be helpful for less experienced surgeons. I recall the first generations looked like lifesaver rings but later became less rigid plates or circles?\n\nRosentreter et al published 2 papers using Ologen to repair the erosion of tube shunt4 or to modulate wound healing and bleb encapsidation after GDD surgery5. However, in the both studies, Ologen implantation was applied several months after the primary surgery. In the second study, the success rate of the Ologen group was significant lower than the controls (only MMC and capsule excision), indicating this collagen graft might affect the post-surgery intraocular pressure. Could authors please share their thoughts about this in the Discussion? This is mostly likely a result of patient cohort under study (i.e. scar formers).\n\nDo the authors use the same postoperative medications with Ologen or can steroids be tapered sooner? Would combining Ologen with a scleral patch make sense in select patients?\n\nCould Ologen improve the IOP if placed around the plate to modify encapsidation?\n\nWhat is the authors’ currently prefered practice pattern for tube shunts, scleral patch or Ologen? Or in other words, how do the authors decide when to use Ologen over a scleral patch graft.",
"responses": []
},
{
"id": "15561",
"date": "09 Aug 2016",
"name": "Arsham Sheybani",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n“The Use of Collagen Matrix As a Patch Graft in Glaucoma” is a well designed study on the use of Ologen as a primary graft in glaucoma shunt surgery. The title is appropriate, though it may benefit from the inclusion of the word “primary” (...”as a primary patch graft”) to distinguish it from other publications on the use of Ologen as a patch graft. The method and analysis are well explained.\n\nAlthough not the primary intent of the study, additional statistical analysis would be useful. This would include reporting the mean change in IOP and medication use post-operatively. While we do not expect that grafts would affect IOP or medication use, data showing IOP reductions could be useful in drawing this conclusion. Reporting the number of previous surgeries on the study eyes and eyelids may elucidate root causes for complications. At present, the supplemental data included only describes surgeries and procedures that occurred after the initial tube placement.\n\nThe conclusions are sensible and justified; however, they do not address whether use of Ologen in any way alters the mechanics of the glaucoma surgery and subsequent IOP management - which again we do not expect. Furthermore, this study describes the use and results of Ologen in one surgeon’s hands. It is possible that results could be different in alternative surgical techniques that may differ in tunnel length and closure.\n\nAs identified by the authors in the discussion, the major limitations of this paper are its small study size and its lack of comparison to other grafts. However it does present a novel method for covering tubes - timely given the shift of glaucoma to a surgical disease. This article serves as a straightforward and succinct study that justifies further investigation and a more robust sub-group analysis.\n\nOverall, this article is in line with the requirements of F1000Research and we support its approval.",
"responses": []
},
{
"id": "19745",
"date": "01 Feb 2017",
"name": "Nathan M. Radcliffe",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors provide a useful report on the practice of replacing a human allograft cornea, sclera, or pericardial graft with porcine collagen matrix (Ologen) to cover the tube entry site during glaucoma drainage device placement. The authors give a reasonable rational for the use of collagen, namely cosmesis, cost and the ability to avoid transplanted human tissue. It should be noted that porcine material is not for everyone, and the use of this material should be discussed with patients prior to placement. The strengths of the study are a reasonable follow up duration and a sufficient number of patients, given that this is the first report of this surgical modification. I say that this number is reasonable, because after 40 cases of a new technique, it is appropriate for the surgeon to pause, carefully analyze the experience, and report the results to the scientific community.\nThe limitations of the study include its retrospective nature and the lack of a control arm. A greater number of patients would include the power of the study, and any future reports on this topic would do well to include more patients. The authors provide adequate support from the literature that a roughly 5% tube erosion rate, as was seen in this study, is the standard.\nAnother limitation of this study relates to the fact that there is a growing body of evidence that glaucoma tube shunt surgery may be safely performed without the use of any patch graft material. Using a long scleral tunnel needle technique, Oscar Albis and colleagues1 reported a 0% erosion rate in 106 Mexican children followed after Ahmed valve placement with no patch graft. The mean follow up was just over two years, no patients were followed for less than six months and some were followed for up to eight years. While this data is encouraging, the reality of the situation is that in the United States, many surgeons are still using patch graft materials, and these materials do have problems, cosmesis and cost being among them. I applaud the authors for making a valuable contribution to the surgical literature and for answering a straightforward but important question: is it reasonable to substitute collagen matrix for other graft materials in glaucoma drainage device surgery? Yes.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1898
|
https://f1000research.com/articles/5-1896/v1
|
01 Aug 16
|
{
"type": "Research Article",
"title": "Moderate chronic fetal alcohol exposure causes a motor learning deficit in adult outbred Swiss-Webster mice",
"authors": [
"Tyler H. Reekes",
"H. Thomas Vinyard III",
"William Echols",
"Andrew J. Eubank III",
"Michael D. Bouldin",
"William H. Murray",
"Stephen Brewer",
"Blake T. Brown",
"Harold L. Willis Jr",
"Zachary Tabrani",
"Carlita B. Favero",
"Erin B.D. Clabough",
"Tyler H. Reekes",
"H. Thomas Vinyard III",
"William Echols",
"Andrew J. Eubank III",
"Michael D. Bouldin",
"William H. Murray",
"Stephen Brewer",
"Blake T. Brown",
"Harold L. Willis Jr",
"Zachary Tabrani",
"Carlita B. Favero"
],
"abstract": "Prenatal ethanol exposure can negatively affect development, causing physical and/or cognitive deficits in the offspring. Behavioral changes are typically characterized during childhood, but they can also persist into adulthood. The extent of Fetal Alcohol Spectrum Disorder (FASD) abnormalities depends upon the amount and manner of ethanol intake, leading to a large variety of animal models. In order to mimic the genetically diverse human condition, we examined an outbred strain of mice exposed to chronic gestational ethanol and characterized subsequent behavioral alterations during adulthood. To detect deficits in cognitive ability and/or motor function, we ran the mice through tests designed to detect either memory/learning ability or motor strength/skill. We tested cognitive responses using the Barnes Maze and the Open Field Aversion Test, and motor skills using Kondziela’s Inverted Screen Test and the rotarod. As adults, the FASD mice showed no significant differences on grip strength, open field, or the Barnes maze; however, we found that outbred mice who had experienced moderate prenatal ethanol exposure were slower to learn the rotarod as adults, though they did not differ in overall performance. Our data suggest a specific FASD vulnerability in motor learning ability, and also open the door to further investigation on the effect of ethanol on brain areas involved in motor learning, including the striatum.",
"keywords": [
"FASD",
"Fetal Alcohol Exposure",
"ethanol",
"chronic",
"outbred",
"Swiss-Webster mice",
"motor skills"
],
"content": "Introduction\n\nEthanol (EtOH) is a toxin that produces variable detrimental physical and neurobehavioral effects in individuals subjected to ethanol exposure in utero, possibly resulting in Fetal Alcohol Spectrum Disorder (FASD). Animal models can replicate the human condition of FASD relatively well (Wilson & Cudd, 2011), but the broad spectrum of human drinking choices made during pregnancy is reflected in the enormous number of different FASD models currently employed by researchers. Even within a single species, differences in acute versus chronic exposure, the timing of the exposure, route of administration, and forced versus voluntary drinking patterns can result in a vast array of possible phenotypes.\n\nThe use of inbred mice has decreased variability in a field with a vast variety of models and made it easier to compare data across paradigms, explore voluntary drinking paradigms, and identify specific genetic differences regulating individual sensitivity to ethanol (Mayfield et al., 2008). Genetic variation causes some strains of mice to voluntarily drink more than other strains (Rhodes et al., 2007) — For example, C57BL/6 mice are heavy drinkers, while DBA2J mice are abstainers (Rhodes et al., 2005). Many FASD studies currently use inbred C57B/6 mice because of the reliability of consumption in this strain (Rhodes et al., 2005).\n\nInbred mouse strains exposed to chronic gestational ethanol have demonstrated deficits in motor function and learning/memory (Brady et al., 2012; Sanchez Vega et al., 2013). Typically, motor reflexes/coordination and learning/memory skills appear to be less dependent on the timing of alcohol exposure than activity and anxiety-related behavioral changes, which have more narrow windows (Mantha et al., 2013).\n\nHowever, use of inbred strains do not faithfully replicate the genetic variation that naturally occurs within the human condition, so outbred strains also provide valuable information about FASD. Outbred Wistar rats are commonly used as a FASD rodent model, but use of outbred mouse models is much less common. Previously reported inbred alcohol-induced deficits may not show up in an outbred mouse model—for example, corticothalamic differences were not observed in an outbred mouse FASD model (White et al., 2015). But a recent paper characterized the effect of moderate prenatal ethanol exposure on outbred Swiss Webster mouse neonate behavior (including surface righting, negative geotaxis, cliff aversion, auditory startle, ear twitch, open field traversal, air righting, and eye opening) and found that ethanol-exposed pups achieved some developmental milestones (surface righting, cliff aversion, and open field traversal) at a different rate than non-exposed control pups (Chi et al., 2016).\n\nTo examine whether observed deficits in outbred mouse neonates can persist into adulthood, we examined the effect of chronic moderate gestational ethanol exposure on adult neurobehavioral outcomes in the offspring previously studied by Chi & colleagues (2016). We measured both learning/memory and motor skill subsets in male mice in order to better understand the behavioral impact of developmental ethanol exposure in an outbred model.\n\n\nMaterials and methods\n\nAll experimentation was compliant with the Hampden-Sydney Institutional Animal Care and Use Committee (IACUC). Twenty-four outbred male Swiss Webster mice were obtained from Ursinus College (four litters per treatment of two-four male mice randomly selected from each litter; 24 animals in total) following completion of the early postnatal testing (Chi et al., 2016). All of the dams in the ethanol treatment group were found to have moderate gestational exposure after using the Drinking in the Dark (Boehm et al., 2008) paradigm, consuming an average of 4.41 g/kg ± .43 standard error (SE) (Table 1). All mice were male and group housed by litter, with a 12-hour light cycle and food and water ad libitum. The weights of the mice at 6 months were not significantly different (by t-test, p=0.43795; mean EtOH=39.16g, SE=0.68; mean H2O=39.46g, SE 1.64). All behavioral tests were performed between 4–6 months of age in the following order to reduce the impact that exposure to testing conditions might have on subsequent testing performance: open field, grip strength, Barnes maze, rotarod (a rotating rod). For each measure, scores for all littermates were averaged together before statistical analysis (12 animals in four litters per treatment group) to account for the variability in gestational ethanol exposure that accompanies a voluntary drinking paradigm.\n\nThe open field test was performed using an opaque box constructed of 6 mm plexiglass (56 cm × 56 cm × 51 cm) as a measure of anxiety-like behavior and locomotion. Testing was modeled after Nunes & colleagues (2011). The first 2 minutes of data were analyzed and the field was divided into nine equal sections. Total squares traveled and time spent in the center square were analyzed for relevance using unpaired t-tests.\n\nThe Barnes maze was used to test for cognitive deficits in learning and memory. Litters were brought into the testing room one at a time to run the Barnes maze using a modified version of the shortened paradigm outlined by Attar et al. (2013) with a 120w light source suspended 1 meter above the stage in the absence of a noxious noise. During habituation and trial days, mice were allowed 30 seconds to enter into the target hole voluntarily before they were coaxed in. In addition, the escape cage remained on the probe day. Entrance into the correct hole, latency to enter, and the percent of holes explored in the opposite quadrant were assessed, as well as the number of holes explored before and after finding the correct hole. Probe day data were analyzed using unpaired t-tests.\n\nThe Kondziela Inverted Screen test was modeled after Deacon (2013) and measures the overall strength of the mice, as well as their ability to maintain their equilibrium while inverted.\n\nThe latency of the mice to fall was tested and an average of their three recordings was taken with an upper limit of 10 minutes per trial. After each trial, the mice were returned to their home cages and allowed to rest approximately 1 hour until the sequential order called for the next test. The mice were randomly selected for the first trial, but the order was maintained for the two subsequent trials. Average latency to fall was analyzed using unpaired t-tests.\n\nThe rotarod was chosen to test both motor learning and endurance. The rotarod (San Diego Instruments) is a spinning rod with dividers separating individual testing spaces. It can be set to turn at a certain rotation per minute (RPM) and to increase RPM at a determined rate. When the mice fall off, sensors on the floor stop the clock and record the number of seconds and the precise RPM.\n\nMice were daily trained on the rotarod over a 3-day period, rested for 1 day, and then exposed to a performance day. All 24 mice were trained to run on the rotarod in a similar manner as Gill & Dietrich (1998) with modifications. Each mouse had a single daily training trial on each of three consecutive days that lasted until they fell off (up to 300 seconds) and animals were not placed on the rod again that day. Animals were rested for one day, followed by a performance assessment as outlined in Nozari et al. (2014). Mice started at zero RPM and accelerated by 6 RPM until the mice fell off. Each mouse ran the performance test three times and the average time per mouse was recorded.\n\nThe order that the mice were selected to run was randomized, but once that order was established, it remained unchanged. The latency to fall on the performance day and the percent improvement from the first day of training to the final performance day were separately analyzed using unpaired t-tests, and performance over the 3-day training period was assessed by a mixed ANOVA using SPSS.\n\n\nResults\n\nIn open field testing, the amount of time spent in the open center square was not significantly different between ethanol-exposed and control animals (by t-test, p=0.2383; mean EtOH=13.72, SE=8.535; mean H2O=6.973, SE=2.476; n=4 litters; Figure 1A). There was no significant difference in locomotor activity between ethanol-exposed and control mice, as measured by the mean number of squares traversed (by t-test, p=0.4231; mean EtOH=34.56, SE=9.691; mean H2O=37.83, SE=7.781; n=4; Figure 1B).\n\n(A) There was no significant difference found between ethanol-exposed mice and control mice in the average number of seconds spent in the center square of the open field over a 2 minute period, as a measure of anxiety-like behavior (by t-test, p=0.2383; mean EtOH=13.72, SE=8.535; mean H2O=6.973, SE=2.476; n=4 litters). (B) No differences were found in the total number of squares crossed in the open field measure, as a measure of locomotion (by t-test, p=0.4231; mean EtOH=34.56, SE=9.691; mean H2O=37.83, SE=7.781).\n\nMice that were exposed to EtOH during development showed no differences on the probe day in total hole exploration (by t-test, p=0.7777; mean EtOH=6.96, SE=1.028; mean H2O=7.67, SE=2.12; n=4 litters), number of incorrect hole explorations before finding the target hole (by t-test, p=0.6009; mean EtOH=5.79, SE=0.3560; mean H2O=6.83, SE=1.766; n = 4), or the number of hole explorations after finding the target (by t-test, p=0.7352; mean EtOH=1.17, SE=0.7876; mean H2O=0.83, SE=0.50; n=4; Figure 2A). There was also no significant difference found in the percentage of holes explored on the opposite quadrant of the maze (by t-test, p=0.1272; mean EtOH=33.33, SE=9.129; mean H2O=57.50, SE=10.13; n= 4; Figure 2B).\n\n(A) There was no significant difference between ethanol-exposed mice and control mice in number of mistakes made before finding the target hole (by t-test, p=0.6009; mean EtOH=5.79, SE=0.3560; mean H2O=6.83, SE=1.766; n = 4 litters), the number of hole explorations after finding the target (by t-test, p=0.7352; mean EtOH=1.17, SE=0.7876; mean H2O=0.83, SE=0.50), or the total hole explorations on the final probe day (by t-test, p=0.7777; mean EtOH=6.96, SE=1.028; mean H2O=7.67, SE=2.12). (B) When the maze was divided into 4 quadrants for analysis, there was also no difference between the ethanol and control mice in the percent of holes located the opposite quadrant that were explored (by t-test, p=0.1272; mean EtOH=33.33, SE=9.129; mean H2O=57.50, SE=10.13).\n\nThere was no significant difference in grip strength between ethanol-exposed and control mice, as measured by the average latency to fall over the three trials (by t-test, p=0.855; mean EtOH=207.396 sec, SE=48.585; mean H2O=193.39, SE=41.221; n=4 litters; Figure 3).\n\nThere was no significant difference found between ethanol-exposed mice and control mice in the latency to fall from the inverted screen as a measure of grip strength (by t-test, p=0.855; mean EtOH=207.396 seconds, SE=48.585; mean H2O=193.39, SE=41.221; n=4 litters).\n\nBy the final performance day, previous ethanol exposure had no effect on the max RPM that mice were able to achieve (by t-test, p=0.399; mean EtOH=7.19, SE=0.68; mean H2O=9.0, SE=1.78; n=4 litters; Figure 4A), indicating that overall motor performance after training was not affected by the gestational ethanol exposure. However, ethanol-exposed animals showed significantly less improvement between the first day of training and the final rotarod performance day, indicating a deficit in the acquisition of this motor skill (by t-test, p=0.008; mean EtOH=30.11% improvement, SE = 2.5%; mean H2O=53.52%, SE=5.5%; Figure 4B).\n\n(A) There was no significant difference between ethanol-exposed mice and control mice in the average max RPM attained on the final performance day (5) (by t-test, p=0.399; mean EtOH=7.19, SE=0.68; mean H2O=9.0, SE=1.78; n=4 litters). (B) On average, control mice improved more than ethanol-exposed mice over the course of the rotarod testing, as measured by the percent change in RPM reached from the first training day (1) to the last performance day (5) (by t-test, p=0.008, EtOH mean=30.11%, SE=2.5%; H2O mean=53.52%, SE=5.5%). (C) We found a possible effect of gestational ethanol exposure on the max RPM at which the mice were able to run over the course of the three training days (F(2,12)=4.812, p=0.029), but this result was not significant after Sidak correction.\n\nBoth saline and ethanol treated mice showed an increase in the latency to fall in succeeding trial days at each age examined, indicating that both groups of mice were learning how to stay on the rod longer. Analysis by mixed ANOVA showed a significant effect of trial day over the three day training period [F(1,6)=37.694; p=0.01; Figure 4C]. We found a possible effect of gestational ethanol exposure on the max RPM at which the mice were able to run over the course of the three training days (F(2,12)=4.812, p=0.029; Figure 4C), but this result was not significant after Sidak correction.\n\n\nDiscussion\n\nOur experiment showed that chronic low/moderate gestational ethanol exposure had no effect on some motor measures in adult outbred mice, including grip strength and overall rotarod performance, yet mice exposed to prenatal ethanol did not improve as quickly when learning the rotarod activity. This adult rotarod learning deficits may be due to long-term ethanol induced changes in specific brain areas that are involved in the acquisition of complex motor skills.\n\nA recent meta-analysis of human FASD concluded that prenatal ethanol exposure can result in gross motor deficits, including gait and balance problems (Lucas et al., 2014). Rodent studies have indicated that ethanol exposure, particularly in the third trimester equivalent, can impact cerebellar Purkinje cell development and alter motor behavior (Maier et al., 1999). The cerebellum is traditionally held responsible for the coordination and execution of motor skills, particularly gait and balance. Ethanol-induced cerebellar damage can cause FASD motor skill deficits, and compromised corticocerebellar circuits may result in impaired visuospatial abilities or other more cognitive processes. However, in rodent models, cerebellar Purkinje cells seem particularly vulnerable to degeneration when exposed to alcohol during the critical period that occurs during the early postnatal period (human third trimester equivalent) (Jaatinen & Rintala, 2008), whereas our mouse model was only exposed during the first and second trimester equivalents.\n\nThe striatum (caudate/putamen) also plays a central role in the acquisition of long term motor skills, and remains of particular interest to alcohol researchers because of its central role in addictive behavior. In rodents, moderate gestational ethanol exposure can alter medium spiny neuron dendritic branching in the striatum pups (Rice et al., 2012), and studies show that developmental ethanol exposure can alter brain-derived neurotrophic factor (BDNF) expression in adult offspring, which can, in turn, impact propensity to drink and lay the groundwork for future addictive behavior, regulated by the striatum (Davis, 2008). BDNF is enriched in the striatum, delivered predominately via corticostriatal afferents, and BDNF polymorphisms have been linked to decreased synaptic plasticity involved in learning both complex and simple learning motor tasks (Cárdenas-Morales et al., 2014). It is possible that BDNF alterations in the striatum could impact synaptic plasticity or neurogenesis and could cause our observed complex motor learning changes.\n\nOur experiment showed that chronic low/moderate gestational ethanol exposure had no effect on open field behavior or Barnes maze performance in outbred adult mice. Our results are in line with some previous studies that have not detected deficits that persist into adulthood for open field, learning/memory tasks, nor locomotor activity in inbred mouse models after chronic gestational ethanol exposure (Boehm et al., 2008; Downing et al., 2009), but it is also possible that our outbred strain introduces enough genetic variability to mask learning or open field phenotypes that have been previously found in inbred strains.\n\nHowever, these ethanol-exposed offspring previously exhibited significantly altered timing to achieve surface righting, cliff aversion, and open field traversal developmental milestones as neonates (Chi et al., 2016), so our results may also indicate some sort of recovery or compensatory response by the central nervous system to the effects of early ethanol exposure has occurred. In most rodent FASD models, motor deficits are visible early in life and have disappeared by adulthood (Patten et al., 2014).\n\nThe value of non-inbred FASD models is currently being explored in other ways: High Alcohol Preferring mice and Low Alcohol Preferring mice have been developed by interbreeding eight inbred strains that display these drinking tendencies (Bice et al., 2011). Testing has shown differences in open field behavior between High and Low Alcohol Preferring mice in the absence of alcohol exposure (Can et al., 2012). Further behavioral characterization of these animals using an FASD model would provide useful motor behavior information, while decreasing the genetic variance present.\n\nFuture ethanol research should pay particular attention to adult phenotypes in order to document the persistence of these changes in FASD models, and to investigate compensatory mechanisms that may allow ethanol-exposed juveniles to overcome these deficits by adulthood. Neuroanatomical analysis should be performed at both juvenile and adult time points to detect possible apoptosis in brain areas involved in motor function. Apart from cell death, both cerebellar and striatal mechanisms that govern the execution of complex motor tasks should be examined, as well as possible long-lasting changes in synaptic plasticity or neurogenesis that may occur following developmental ethanol exposure.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data of motor learning deficit in adult FASD mice, 10.5256/f1000research.9237.d130218 (Reekes et al., 2016).",
"appendix": "Author contributions\n\n\n\nEC and all other authors conceived the study, designed the experiments, prepared all manuscript drafts, and have agreed to the final content. SB, MB, WE, TR, BB, and HW carried out the open field and Barnes maze experiments. TV, AE, ZT, and WM carried out the motor testing. All authors were involved in the revision of the draft manuscript. CF performed the gestational ethanol exposure and acted as a consultant.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nA portion of this work was funded by a San Diego Instruments Rotarod Equipment Research Loan Award from the Faculty for Undergraduate Neuroscience to EC.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to thank John Reekes for logistical assistance and Jennie Jenkins for laboratory support.\n\n\nReferences\n\nAttar A, Liu T, Chan WT, et al.: A shortened Barnes maze protocol reveals memory deficits at 4-months of age in the triple-transgenic mouse model of Alzheimer's disease. PLoS One. 2013; 8(11): e80355. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBice PJ, Lai D, Zhang L, et al.: Fine mapping quantitative trait loci that influence alcohol preference behavior in the High and Low Alcohol Preferring (HAP and LAP) mice. Behav Genet. 2011; 41(4): 565–570. PubMed Abstract | Publisher Full Text\n\nBoehm SL 2nd, Moore EM, Walsh CD, et al.: Using drinking in the dark to model prenatal binge-like exposure to ethanol in C57BL/6J mice. Dev Psychobiol. 2008; 50(6): 566–578. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrady ML, Allan AM, Caldwell KK: A limited access mouse model of prenatal alcohol exposure that produces long-lasting deficits in hippocampal-dependent learning and memory. Alcohol Clin Exp Res. 2012; 36(3): 457–466. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCan A, Grahame NJ, Gould TD: Affect-related behaviors in mice selectively bred for high and low voluntary alcohol consumption. Behav Genet. 2012; 42(2): 313–322. PubMed Abstract | Publisher Full Text\n\nCárdenas-Morales L, Grön G, Sim EJ, et al.: Neural activation in humans during a simple motor task differs between BDNF polymorphisms. PLoS One. 2014; 9(5): e96722. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChi P, Aras R, Martin K, et al.: Using Swiss Webster mice to model Fetal Alcohol Spectrum Disorders (FASD): An analysis of multilevel time-to-event data through mixed-effects Cox proportional hazards models. Behav Brain Res. 2016; 305: 1–7. PubMed Abstract | Publisher Full Text\n\nDavis MI: Ethanol-BDNF interactions: still more questions than answers. Pharmacol Ther. 2008; 118(1): 36–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeacon RM: Measuring the strength of mice. J Vis Exp. 2013; (76): e2610–e2610. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDowning C, Balderrama-Durbin C, Hayes J, et al.: No effect of prenatal alcohol exposure on activity in three inbred strains of mice. Alcohol Alcohol. 2009; 44(1): 25–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGill K, Deitrich RA: Acute tolerance to the ataxic effects of ethanol in short-sleep (SS) and long-sleep (LS) mice. Psychopharmacology (Berl). 1998; 136(1): 91–98. PubMed Abstract | Publisher Full Text\n\nJaatinen P, Rintala J: Mechanisms of ethanol-induced degeneration in the developing, mature, and aging cerebellum. Cerebellum. 2008; 7(3): 332–347. PubMed Abstract | Publisher Full Text\n\nLucas BR, Latimer J, Pinto RZ, et al.: Gross motor deficits in children prenatally exposed to alcohol: a meta-analysis. Pediatrics. 2014; 134(1): e192–209. PubMed Abstract | Publisher Full Text\n\nMaier SE, Miller JA, Blackwell JM, et al.: Fetal alcohol exposure and temporal vulnerability: regional differences in cell loss as a function of the timing of binge-like alcohol exposure during brain development. Alcohol Clin Exp Res. 1999; 23(4): 726–734. PubMed Abstract | Publisher Full Text\n\nMantha K, Kleiber M, Singh S, et al.: Neurodevelopmental timing of ethanol exposure may contribute to observed heterogeneity of behavioral deficits in a mouse model of fetal alcohol spectrum disorder (FASD). J Behav Brain Sci. 2013; 3(1): 85–99. Publisher Full Text\n\nMayfield RD, Harris RA, Schuckit MA: Genetic factors influencing alcohol dependence. Br J Pharmacol. 2008; 154(2): 275–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNozari M, Shabani M, Hadadi M, et al.: Enriched environment prevents cognitive and motor deficits associated with postnatal MK-801 treatment. Psychopharmacology (Berl). 2014; 231(22): 4361–4370. PubMed Abstract | Publisher Full Text\n\nNunes F, Ferreira-Rosa K, Pereira Mdos S, et al.: Acute administration of vinpocetine, a phosphodiesterase type 1 inhibitor, ameliorates hyperactivity in a mice model of fetal alcohol spectrum disorder. Drug Alcohol Depend. 2011; 119(1–2): 81–87. PubMed Abstract | Publisher Full Text\n\nPatten AR, Fontaine CJ, Christie BR: A comparison of the different animal models of fetal alcohol spectrum disorders and their use in studying complex behaviors. Front Pediatr. 2014; 2: 93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRice JP, Suggs LE, Lusk AV, et al.: Effects of exposure to moderate levels of ethanol during prenatal brain development on dendritic length, branching, and spine density in the nucleus accumbens and dorsal striatum of adult rats. Alcohol. 2012; 46(6): 577–584. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReekes TH, Vinyard HT III, Echols W, et al.: Dataset 1 in: Moderate Chronic Fetal Alcohol Exposure Causes a Motor Learning Deficit in Adult Outbred Swiss-Webster Mice. F1000Research. 2016. Data Source\n\nRhodes JS, Best K, Belknap JK, et al.: Evaluation of a simple model of ethanol drinking to intoxication in C57BL/6J mice. Physiol Behav. 2005; 84(1): 53–63. PubMed Abstract | Publisher Full Text\n\nRhodes JS, Ford MM, Yu CH, et al.: Mouse inbred strain differences in ethanol drinking to intoxication. Genes Brain Behav. 2007; 6(1): 1–18. PubMed Abstract | Publisher Full Text\n\nSanchez Vega MC, Chong S, Burne TH: Early gestational exposure to moderate concentrations of ethanol alters adult behaviour in C57BL/6J mice. Behav Brain Res. 2013; 252: 326–333. PubMed Abstract | Publisher Full Text\n\nWhite SA, Weber JN, Howard CD, et al.: Effects of binge ethanol exposure during first-trimester equivalent on corticothalamic neurons in Swiss Webster outbred mice. Neuroreport. 2015; 26(18): 1083–1088. PubMed Abstract\n\nWilson SE, Cudd TA: Focus on: the use of animal models for the study of fetal alcohol spectrum disorders. Alcohol Res Health. 2011; 34(1): 92–8. PubMed Abstract | Free Full Text"
}
|
[
{
"id": "15345",
"date": "16 Aug 2016",
"name": "Derek Hamilton",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper is motivated in part by the discrepancy between genetic variation in humans that is absent in many animal models of FASD that involve assessments in inbred strains. The authors examined learning/memory and motor strength/skill in adult outbred (Swiss-Webster) mice exposed to gestational alcohol exposure (first two trimester equivalent). Exposed mice showed no statistically significant deficits in grip strength, open field exploration, or in the Barnes Maze. Exposed mice were slower to learn the rotorod, but did not differ in overall performance. The authors conclude that motor learning abilities are affected by moderate gestational ethanol exposure in the absence of conspicuous effects on motor performance. The strengths of the study are 1) the use of multiple behavioral and cognitive tasks within subjects, 2) measurements taken during adulthood to address persistence of effects, 3) the use of an established voluntary drinking paradigm, and 4) potential for motivating more research on genetic variability and ethanol exposure. Addressing several points outlined below would strengthen the paper and its impact on the field.\n\nFor analyses the litter was used as the unit of analysis. Litter effects are commonly addressed by limiting the number of animals from a given litter used in a particular study. In this report, unequal numbers (2-4 animals) from 4 litters were used. Thus, the n for each group was 4. Considering this, some effects that are numerically noteworthy but not statistically significant (e.g., open field, time in center, Fig 1A; Barnes maze exploration of holes in the opposite quadrant, Fig 2B; final rotorod performance, Fig 4A) could be underpowered. Estimates of effect sizes and power calculations should be provided. This issue takes on added importance given that one of the principal conclusions concerns the lack of motor performance deficits following prenatal alcohol exposure.\n\nAn alternative analysis could also be considered. Though litters are commonly used as the unit of analysis in developmental studies with proper justification and design, averaging data from multiple animals in a given litter also eliminates potentially important individual variation. An analysis treating animals as nested within litters within groups (Lazic and Essioux, 2013) represents an alternative that could yield different statistical outcomes.\n\nThe relative sensitivity of the tests and measures should be considered. More fine-grained behavioral assessments may be more sensitive to ethanol exposure. Further, different motor behaviors that are distinguished by effector systems and neural circuitry (e.g., orofacial movements, skilled reaching; see e.g., Hamilton, 2014) will likely yield distinct results, which should be considered in integrating the present results and conclusions with the extant literature.\n\nOne of the primary motivations noted in the paper is the issue of using inbred vs. outbred strains in the study of fetal alcohol effects. The present study does not directly compare outbred and inbred strains statistically. The potentially large variation in parameters of ethanol exposure models and behavioral methods make drawing conclusions from qualitative comparisons across studies difficult. The authors should consider providing some recommendations for how to more fully address the issue of genetic variation and fetal alcohol effects.\n\nThe authors note that inbred mouse strains display deficits in motor function and learning/memory. It is important to contextualize this discussion with reference to critical variables including the timing, dose, duration, and route of ethanol administration. Further, an expanded treatment of the motor effects following prenatal alcohol exposure would help place the current findings in the broader landscape of the literature.\n\nThe authors should provide estimates of blood ethanol concentrations associated with the level of drinking, and clarify what they consider to be moderate ethanol exposure.\n\nThe raw data files contain some discrepancies or omissions that should be corrected. Raw data 1 (open field) contains litter numbers that do not correspond to same mouse numbers as provided in other files (e.g., .Raw data 3). Raw data 2 (Barnes maze) only includes litter numbers but not individual mouse id values.\n\nMINOR ISSUES\n\nAbstract : “Mice who” should be “mice that”\n\nThe introduction should provide more information and citations to relevant articles regarding various ethanol exposure paradigms.\n\nThe second sentence of the third introduction paragraph is a bit vague.\n\nAs written the second sentence of the fifth introduction paragraph reads as if the use of male mice is critical to advancing understanding of the behavioral consequences of ethanol exposure.\n\nFor the open field please include the lighting parameters.\n\nThe methods for the Barnes maze indicate that the task was conducted with a light source but no auditory stimulus. This is inconsistent with the methods cited (Attar et al. 2013).\n\nPage 7. In the sentence beginning “In rodents, moderate gestational ethanol exposure can alter medium spiny neuron dendritic branching in the striatum pups (Rice et al., 2012), …” The word “pups” should be removed.",
"responses": []
},
{
"id": "16659",
"date": "28 Sep 2016",
"name": "Kelly J. Huffman",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study examines the long-term behavior effects of prenatal ethanol exposure in an outbred murine stock. The authors discuss a very important idea that the sole use of inbred mice to study FASD is problematic due to the lack of genetic variation. If scientists attempt to model a human condition like FASD, it is ultimately more informative to use a genetic model that is more similar to humans, who are by nature an outbred stock.\n\nIssues: The NIH has stated that scientist should no longer neglect female mice as subjects in studies like this and if the authors do not wish to include females in their analyses (if they did it would strengthen the paper) then they should give justification in the text for the sole use of males. Sex could be used a potential covariate, as is time of measurement. The 4-6 month age time period may cause some differences in the data. For example, a 6-month old mouse has had 50% more time for neural plasticity than mice that are 4 months old. I would not require redoing the experiments but perhaps including age as a covariate to determine whether it plays a role in the effects would be a wise idea. There are two reports left out of the discussion that should be mentioned as they employ use of an outbread stock of mice (CD1) in a prenatal alcohol exposure model. They should be includes in the discussion (El Shawa et al., 2013 and Abbott et al., 2016).\nIn summary, the title is appropriate for the article and the abstract represents a suitable summary of the work. The design, methods and analysis of the results from the study are appropriate and explained well. I think that the conclusions drawn by the authors draw are sensible, balanced and justified. All the data provided is in a usable format/ structure, and enough information has been provided so that the experiment can be replicated.\n\nOverall this is a good paper.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1896
|
https://f1000research.com/articles/5-1890/v1
|
29 Jul 16
|
{
"type": "Research Note",
"title": "The frequency distribution of vitamin D Receptor fok I gene polymorphism among Ugandan pulmonary TB patients",
"authors": [
"Ester L. Acen",
"William Worodria",
"Peter Mulamba",
"Andrew Kambugu",
"Joseph Erume",
"William Worodria",
"Peter Mulamba",
"Andrew Kambugu",
"Joseph Erume"
],
"abstract": "Background: Mycobacterium tuberculosis (TB) is still a major problem globally and especially in Africa. Vitamin D deficiency has been linked to TB in the past and studies have found vitamin D deficiency to be common among Ugandan TB patients. The functional activity of vitamin D is dependent on the genotype of the vitamin D receptor (VDR) polymorphic genes. Recent findings have indicated that VDR polymorphisms may cause increased resistance or susceptibility to TB. The vitamin D ligand and its receptor play a pivotal role in innate immunity by eliciting antimicrobial activity, which is important in prevention of TB. The fok I vitamin D receptor gene has extensively been examined in TB patients but findings so far have been inconclusive. Objectives: This study sought to investigate the frequency distribution of the VDR fok I gene polymorphisms in pulmonary TB patients and controls. Methods: A pilot case control study of 41 newly diagnosed TB patients and 41 healthy workers was set up. Vitamin D receptor fok I gene was genotyped. Results: The frequency distribution of fok I genotype in Ugandan TB patients was 87.8% homozygous-dominant (FF), 7.3% (Ff) heterozygous and 4.8% (ff) homozygous recessive. For normal healthy subjects the frequencies were (FF) 92.6%, (Ff) 2.4% and (ff) 4.8%. No significant difference was observed in the FF and ff genotypes among TB patients and controls. The Ff heterozygous genotype distribution appeared more in TB patients than in controls. A significant difference was observed in the fok I genotype among gender p value 0.02. No significant difference was observed in ethnicity, p value 0.30. Conclusions: The heterozygous Ff fok I genotype may be associated with TB in the Ugandan population.",
"keywords": [
"Vitamin D Receptor",
"Polymorphism",
"Fok I genotypes",
"Tuberculosis",
"Uganda Introduction"
],
"content": "Introduction\n\nAccording to the World Health Organization (WHO) report on the ‘Use of high burden country lists for TB by WHO in the post-2015 era (WHO/HTM/TB/2015.29), Uganda was removed from the 22 high TB burden countries. However Uganda is still among the 41 high TB/human immunodeficiency virus (HIV) burden countries1 and TB remains a major public health problem. Although it is still unclear how individuals develop active TB2 there is emerging evidence of multi drug resistant TB (MDRTB). Genetic predisposition to TB has been suggested in several studies3,4 but the immunological associations with genetic polymorphisms are still unclear5. Vitamin D insufficiency is common worldwide and recent studies have shown that vitamin D insufficiency is associated with a higher risk of active TB. The vitamin D receptor gene (VDR) has been identified as a candidate gene for TB susceptibility, however; studies reporting from different ethnic groups have been inconsistent6. Various VDR polymorphisms have been identified and associated with TB susceptibility or resistance2,7,8. Similarly the widely studied and functional single nucleotide polymorphism (SNP) rs2228570 of VDR fok I gene has shown variations. This polymorphism is due to the alteration in one of the start codons in which the amino acid thymine is replaced with cytosine (T/C). The dominant homozygous FF variant has high transcriptional activity with three amino acids less since translation starts at the second codon, while the homozygous recessive (ff) has 427 amino acids and is the longer form. The absence or presence of the restriction site is designated as F or f respectively7,11. This study investigated the frequency distribution of VDR fok I gene among TB patients and compared it with controls in a Ugandan population.\n\n\nMethods\n\nAfter obtaining permission and informed consent (see ethical considerations) a pilot study in newly diagnosed smear positive TB patients and healthy controls was conducted between the months of April and June 2013 at the Mulago National Referral and Teaching Hospital located in North of Kampala, Uganda. By consecutive sampling adult patients who presented with persistent cough for more than 3 weeks had a positive Ziehl-Nielsen smear test for the first time and signed a written consent were considered eligible. Health workers and medical trainees who worked at the TB out patient’s clinic and other wards exposed to TB were matched for sex and age with the control group. All subjects were screened for human immunodeficiency virus (HIV) and data and consent forms were filled. Blood was then collected into tubes with ethylenediaminetetraacetic acid (EDTA) anticoagulant (Becton, Dickinson and company New Jersey USA) and whole blood samples were stored at 2–8°C before DNA column extraction.\n\nGenotyping of VDR fok I gene was performed by PCR–direct sequencing method on ABI 310 sequencer. Human genomic DNA was extracted using the Genotype DNA isolation kit version 2.0 (HAIN Life Science, Germany) according to the manufacturer’s instructions. To check the purity genomic DNA was run on a 1.0% agarose gel in a 1 × Tris Acetate EDTA (TAE) buffer containing ethidium bromide for 1 hour at 120 volts and visualized under an UV transilluminator. The primer sequences used in our study were forward A 5’-GC TGG CCC TGG CAC TGA CTC TG TCT -3’ and reverse 5’-ATG GAA ACA CCT TGC TTC TTC TCC CTC as -3’ described by Harris et al. (1997). PCR was performed in 15.1 µl reaction volumes which contained 7 µl of PCR water, 1 µl of 25 mM Mgcl2, 1 µl of 10X master mix (Fisher biotec company Australia), 1 µl of 0.22 mg of each of the forward and reverse primers (Integrated DNA technologies company USA), 0.1 µl of 5U of Pre heated Taq polymerase and finally 5 µl of pure DNA were added and the reaction was thoroughly mixed. The PCR GTQ Thermocycler programme involved an initial denaturation at 95°C for 5 minutes, followed by 30 cycles 70°C annealing for 45 seconds, elongation at 72°C for 1 minute and a final elongation at 72°C, for 10 minutes. Amplified DNA was purified using an extraction kit (JET quick company, USA) according to manufacturer’s instructions. Cycle sequencing PCR was performed using the Big Dye terminator KIT version 3, in the Thermocycler Gene Amp PCR system 9700. The cycling programme involved denaturation at 96°C for 1 minute for 1 cycle ,annealing at 96°C for 10 minutes 30 cycles, elongation at 70°C for 30 seconds and a final elongation at 60°C for 5 minutes. The Dye EX 2.0 spin kit (250) (QIAGEN, Germany) was used to remove dye and other PCR products following the manufacturer’s instructions. Sequencing of the PCR products was done using the ABI Big Dye Termination kit (Applied Biosystems, USA) and the ABI prism 310 Genetic analyzer (Applied Biosystems). Sequences obtained were compared to those in the Gene Bank database by applying BLAST; NCBI tool. Finally DNA baser sequence assembly software version 4.7.0 bioinformatics tool was used for the Identification of fok I gene polymorphism. Polymorphism of FF genotype was identified on sequences with a start codon of ACG instead of an ATG followed by an ATG downstream. The homozygous ff had a start codon of ATG and consequently another ATG in the sequence. The heterozygous Ff genotype had an ACG, ACG and an ATG.\n\nThe sample size was calculated based on data of a previous study9 with mean of 78.3 for TB and 83.5 for healthy controls and a total of 82 samples at a power of 80% would detect an effect. We therefore divided the samples into cases and controls for a pilot study. Data were summarized into odds ratios, 95% confidence interval (95% CI), and alpha of p<0.05 was considered significant using STATA software (Stata Corp. STATA 12.0, College Station, Texas, USA). The frequency of vitamin D fok I gene was determined using percentages. A Chi square test was used to determine the allele and genotype frequency distribution and potential deviation from Hardy Weinberg equilibrium. Fok I genotypes were categorized as FF for dominant, Ff heterozygous and ff for homozygous variant.\n\n\nResults\n\nA total of 82 participants were enrolled in the study. Of these, 41 had pulmonary TB while 41 were healthy subjects with no history of the disease. Majority of participants were males (66.0%). Among the participants with TB 73.2% were males while the females were 26.8%. The percentage of HIV/TB patients in this study of 24.4% was lower than the non HIV/TB patients. Thirteen (15.9%) of the 82 participants were HIV seropositive, of these three were from the control group. Most of the TB patients were drivers, farmers, teachers and others. Other lifestyle factors did not show significant variation. Details of the above results are shown in Table 1. Only four (10%) participants were aware of exposure to TB (data not shown).\n\nFollowing DNA extraction and amplification of the fok I gene in all 82 samples the electrophoresis results showed a band between 200 and 300 bps (Figure 1). The PCR products were then sequenced and a BLAST on the query sequences was done against those in the NCBI Gene bank to confirm the sequences of fok I. The gene product was named vitamin D3 receptor isoform VDRB1 vitamin D3 receptor isoform VDRA. It showed sequences between 238 bp and 260 bp with a percentage identity between 98–99% (see Dataset 1).\n\nLanes: M=100bp DNA Ladder, 1=Positive Control, 2=Negative Control, 3–7 and 8–13 = Representative sequencing PCR products from controls and cases respectively.\n\nThe frequencies of Fok I genotypes, FF, Ff, and ff in TB patients and in the controls were assessed. Table 2 shows the genotype and allele frequency distribution of this study population with their odds ratio, confidence intervals and p values. No significant difference was observed in the distribution of the homozygous genotypes in the study population and the two alleles. FF genotype was used as the reference and other genotypes were compared while the f allele was used as reference to the F allele (Table 2). The heterozygous Ff genotype was associated with TB, (OR 3.2) since it occurred more in the TB patients than the healthy subjects as shown in Table 2. There was a predominance in the allele distribution of F dominant verse the f recessive allele in both the TB and control group. Both HIV and non HIV individuals predominantly had the FF genotype with only one HIV/TB patient having the recessive ff genotype. Hardy Weinberg equilibrium was tested to establish if the genotype and allele distribution among the TB patients and the control population was in equilibrium. The Chi square test estimated a deviation coefficient of 0.041 in the TB patients and 0.045 in the healthy subjects.\n\nOR =Odds ratio, CI= confidence interval, FF reference genotype and f reference allele\n\nAn analysis of fok I gene was performed to determine the frequency distribution of the fok I gene polymorphism among male and female subjects (Table 3). The frequency distribution of fok I FF homozygous dominant genotype was male was 57.3% for male and 32.9 % for females. There was one female with heterozygous Ff genotype while homozygous recessive ff genotype was only found male subjects. These results showed a significant difference in the distribution of the fok I genotypes among male and female subjects given a p value of 0.02. When a multivariate analysis was further performed against fok I, gender and other variables a p 0.01 was obtained.\n\n\nDiscussion\n\nVDR polymorphisms have been studied in different populations and have shown various results. We document for the first time the frequency distribution of VDR polymorphism in the Ugandan population. This study reports a frequency of 87.8% of FF genotype in the Pulmonary Tuberculosis (PTB) patients; however previous studies have reported a 65% frequency of FF genotype among Indian PTB patients, 72% among the Venda of South Africa and 62% among west-Africans. The low frequency of f allele in this population is consistent with another study10 showing that the f allele of Fok I occurs less frequently in Africans compared to Caucasians and Asians, (Africans 24%, Caucasians 34%, and Asians 51%). The ff genotype frequency of 4.8% was similar to the one described for PTB patients in the African case control study of Gambia Guinea and Guinea Bissau. The ff genotype was reported to be 5% in an Indian healthy population11, while it was 6% in the African American population and 3% in the Venda of South Africa12. The ff genotype was only found among the males in this study. This observation differs from an Indian study showing the FF genotype only in males and a significant difference in the genotypic distribution between males and females. The VDR polymorphism was not in equilibrium with Hardy Weinberg and this could be attributed to the low frequency of the ff genotype. Among other studies that have also reported disequilibrium in the VDR are the studies in the three West African countries6 and a study on the Paraguay population8. Disequilibrium observed in these studies may be due to genotyping errors but was overcome by use of high quality DNA which was achieved by secondary PCR. Other factors are genetic drift and mutations. There was no significant difference in the distribution of fok I genotype in different ethnic groups of our study population, (p value 0.30). However a difference was observed when the analysis was based on gender. The low numbers of females in this study may probably be explained by the female sex being a protection against TB according to a study13. Among the 5 % to 10 % of individuals who develop active disease about 70% are males.\n\n\nConclusion\n\nAlthough the frequency distribution of the homozygous fok I genotypes was not significantly different in the Ugandan TB patients and controls the Ff heterozygous genotype appears to be associated with TB in the Ugandan population. Therefore further studies are needed to elucidate this relationship.\n\n\nData availability\n\nF1000Research: Dataset 1. Human VDR fok I gene nucleotide sequences, 10.5256/f1000research.9109.d13042014\n\n\nConsent\n\nWritten informed consent to participate in the study and publish these data was obtained from all participants.\n\n\nEthical considerations\n\nPermission to carry out the study was obtained from the Research and Ethics Committee Mulago Hospital, (ref MREC: 329), the Institutional Review Board of the School of Biomedical Sciences Higher Degrees Research and Ethics committee, (ref SBS 108), and from the Uganda National Council of Science and Technology (ref No.1431). Before investigations, informed consent was obtained from the study subjects. Patients’ personal information was kept confidential by using serial codes instead of names on the questionnaire.",
"appendix": "Author contributions\n\n\n\nAE: conceived the study, AE and EJ: designed the experiments, AE, WW and EJ: developed proposal and was involved in research, MP contributed to study design and other statistics expertise AE WW, MP, KA and EJ prepared the first draft of manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed\n\n\nGrant information\n\nThis research was sponsored by the Makerere University and the Swedish International Development Cooperation Agency Bilateral under the Gender Main streaming Directorate. Support for manuscript writing was provided by Fogarty International Center, National Institutes of Health (grant # D43TW009771 “HIV co-infections in Uganda: TB, Cryptococcus, and Viral Hepatitis.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nWe would like to acknowledge the study participants, Mulago hospital and MBN laboratories for analysis.\n\nWe thank the Swedish International Development Cooperation Agency through Gender Main streaming, Makerere University for funding this research.\n\n\nReferences\n\nWHO: Use of high burden country lists for TB by WHO in the post-2015 era Global report 2015. Geneva: World Health Organization; 2015. Reference Source\n\nSelvaraj P, Narayanan PR, Reetha AM: Association of vitamin D receptor genotypes with the susceptibility to pulmonary tuberculosis in female patients & resistance in female contacts. Indian J Med Res. 2000; 111: 172–179. PubMed Abstract\n\nBellamy R, Ruwende C, Corrah T, et al.: Tuberculosis and chronic hepatitis B virus infection in Africans and variation in the vitamin D receptor gene. J Infect Dis. 1999; 179(3): 721–724. PubMed Abstract | Publisher Full Text\n\nComstock GW: Frost revisited: the modern epidemiology of tuberculosis. Am J Epidemiol. 1975; 101(5): 363–82. PubMed Abstract\n\nBerrington WR, Hawn TR: Mycobacterium tuberculosis, macrophages, and the innate immune response: does common variation matter? Immunol Rev. 2007; 219(1): 167–186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBornman L, Campbell SJ, Fielding K, et al.: Vitamin D receptor polymorphisms and susceptibility to tuberculosis in West Africa: a case-control and family study. J Infect Dis. 2004; 190(9): 1631–41. PubMed Abstract | Publisher Full Text\n\nGross C, Krishnan AV, Malloy PJ, et al.: The vitamin D receptor gene start codon polymorphism: a functional analysis of FokI variants. J Bone Miner Res. 1998; 13(11): 1691–9. PubMed Abstract | Publisher Full Text\n\nWilbur AK, Kubatko LS, Hurtado AM, et al.: Vitamin D receptor gene polymorphisms and susceptibility to M. tuberculosis in native Paraguayans. Tuberculosis (Edinb). 2007; 87(4): 329–337. PubMed Abstract | Publisher Full Text\n\nWejse C, Olesen R, Rabna P, et al.: Serum 25-hydroxyvitamin D in a West African population of tuberculosis patients and unmatched healthy controls. Am J Clin Nutr. 2007; 86(5): 1376–83. PubMed Abstract\n\nUitterlinden AG, Fang Y, Van Meurs JB, et al.: Genetics and biology of vitamin D receptor polymorphisms. Gene. 2004; 338(2): 143–56. PubMed Abstract | Publisher Full Text\n\nBhanushali AA, Lajpal N, Kulkarni SS, et al.: Frequency of fokI and taqI polymorphism of vitamin D receptor gene in Indian population and its association with 25-hydroxyvitamin D levels. Indian J Hum Genet. 2009; 15(3): 108–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVetten M: Polymorphisms in the regulatory region of the Vitamin D Receptor gene (VDR): In silico analysis, tuberculosis association and functional impact. 2009. Reference Source\n\nLienhardt C, Bennett S, Del Prete G, et al.: Investigation of environmental and host-related risk factors for tuberculosis in Africa. I. Methodological aspects of a combined design. Am J Epidemiol. 2002; 155(11): 1066–1073. PubMed Abstract | Publisher Full Text\n\nAcen E: Dataset 1 in: The frequency distribution of Vitamin D Receptor fok I Gene polymorphism among Ugandan pulmonary TB patients. F1000Research. 2016. Data Source"
}
|
[
{
"id": "15525",
"date": "19 Aug 2016",
"name": "Simon H. Martin",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAcen et al. investigated whether the fok I polymorphism in the vitamin D receptor gene is associated with tuberculosis in 82 (41 infected and 41 uninfected) Ugandans.\nThis was a pilot study with a fairly small sample size, and I have evaluated it as such. In general, I see no fundamental flaws in the study design. However, I have some serious concerns about the statistical analyses and interpretation thereof, as well as several minor concerns, as detailed below.\nMajor concerns\nAssociation of heterozygous Ff with TB\nIn the Results it is claimed that \"The heterozygous Ff genotype was associated with TB, (OR 3.2) since it occurred more in the TB patients than the healthy subjects. The Abstract also states that \"The Ff heterozygous genotype distribution appeared more in TB patients than in controls\". And the discussion says \"the Ff heterozygous genotype appears to be associated with TB in the Ugandan population\". I think these claims are misleading, as the result was not statistically significant at all. In fact, when you consider the raw numbers (3 of the 41 TB patients were Ff and 1 of the 41 healthy samples were Ff), I think it is clear that there is no evidence in this data for an association between Ff and TB status. I feel that the text should be revised to acknowledge the lack of statistical support for this claim.\n\nAssociation between genotype and gender\nThe section \"Genotype fok I distribution by gender\" and Table 3 is not very clear. It appears to state that genotype FF occurred at a frequency of 57.3% in males and 32.9% in females, but this seems to be a mistake. Table 3 shows that FF occurred in 47 of 54 males (87%) and 27 of 28 females (96.4%). There is no information about the statistical test performed, but the genotype frequencies look to me to be too similar to justify the p-value of 0.02 in Table 3. There is also no information given about the multivariate analysis mentioned. Therefore, I recommend that this entire section needs to be revised, with clearer descriptions of the statistical analyses and correction of the genotype frequencies.\n\nDeviation from Hardy-Weinberg Equilibrium\nMore details of this chi-square test are required. As far as I can tell, because the frequency of the f allele is so low, the expected number of ff samples according to HWE is zero, so I'm not sure how the chi squared test could be performed. There are methods to test for a deviation from HWE that don't have the same problem with rare alleles. For example, see Wigginton et al. (2005).\n\nMinor concerns\nAbstract line 1 – Mycobacterium tuberculosis is not TB, it is the causative agent of TB.\n\nIntroduction Par 1 - “Similarly the widely studied and functional single nucleotide polymorphism (SNP) rs2228570 of VDR fok I gene has shown variations.” The meaning of this sentence is not clear.\n\nDataset 1 only includes 35 sequences. Where are the others?\n\nReferences missing in a few places: - Introduction Par 1 - “Vitamin D insufficiency is common worldwide and recent studies have shown that vitamin D insufficiency is associated with a higher risk of active TB” - Discussion Par 1 - “The ff genotype frequency of 4.8% was similar to the one described for PTB patients in the African case control study of Gambia Guinea and Guinea Bissau.”",
"responses": []
},
{
"id": "15526",
"date": "31 Aug 2016",
"name": "Lori B. Chibnik",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes a case-control study carried out at the Mulago National Referral and Teaching Hospital to assess the possible association between the fok I gene and TB diagnosis. The cases were defined as newly diagnosed cases of TB and were matched to TB negative controls taken from health workers at the hospital. The minor allele frequency (MAF) was smaller than expected with the overall MAF = 7%. While some differences were found between TB cases and controls, none were statistically significant. It is a fairly straight forward analysis and Dr. Acen and colleagues made a valiant effort to explore this question. Some comments:\nThe controls were chosen from health workers and trainees at the hospital (age and sex matched). This brings in some confounding as health workers and medial trainees tend to be different than the general population in terms of education, SES, etc. some of which include risk factors for TB. This needs to be addressed in the discussion as a limitation.\n\nAlso, were the controls tested to confirm they do not have TB, this should also be mentioned.\n\nI am assuming the health workers are the controls, the wording is a bit strange. The manuscript needs to be proof-read by a native or more fluent English speaker\n\nThe authors state that cases and controls were matched on sex, but the numbers (table 1) do not show this. There are 30 males in the TB group and only 24 in the controls.\n\nThe main result of this paper is the OR of 3.2 presented in table 2. The authors claim that this is a real effect and shows a possible relationship between TB and the fok I gene. They neglect to mention the 95% CI that includes the null (0.3 – 31.8) in text which is misleading. With such small numbers and a non-significant effect this is not a meaningful finding. It argues that more research is needed, but it is hard to argue that there is a real effect shown in these data.\n\nStatistical Analysis section and Table 2:\n\nThe authors list numbers (78.3 and 83.5) as means for use in a power calculation. What do these numbers stand for? If they are allele frequencies then that should be stated.\n\nWith such low frequencies in the some of the cells, a chi-square test is not appropriate. A Fisher’s exact test should be used to determine p-values\n\nWhy is FF used as a reference genotype but f used as reference allele? This makes the interpretation a bit more confusing.\n\nHow were the p-values calculated for Table 2? This needs to be indicated. What is the p-value for the reference groups from?\n\nTable 3\n\nThe analyses based on sex needs to be described in the statistical analysis section.\n\nWhat was the impetus to compare frequencies based on sex? It seemed to come out of nowhere.\n\nThe authors should focus more on the differences in MAF for fok I gene in their population compared to other published populations of different ancestry.",
"responses": []
},
{
"id": "16253",
"date": "13 Sep 2016",
"name": "Range Nyagosya",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe vitamin D receptor plays an important role in the immune system and the roles have widely been studied. Several studies have been conducted to assess the frequency distribution as well as roles of Vitamin D’s in various diseases conditions; including; multiple sclerosis, cancer, diabetes and TB. However, the findings have been varying and inconclusive, imposing the need of carrying further investigations in different populations and settings to determine the roles and frequency distrition. It was on this varying state of Vitamin D that the authors, Ester et al, conducted a study in Uganda aiming to investigate the frequency distribution of the VDR gene (fok I) polymorphisms in pulmonary TB patients compared to controls in Ugandan population.\nThe article though short and with relatively small sample size of 82 (TB patients and health controls) is well written, appropriate description is given covering all sections; introduction, objectives methods, results, discussions and conclusion. The study found no significance difference in the frequency distribution of homozygous –dominant (FF) and homozygous receive (ff) genotype among TB patients and controls. However, heterozygous (Ff) genotype was more frequent among TB patients compared to controls (7.3% vs 2.4%). The author concluded that Ff gene may be associated with TB because it occurred more in TB patients than in health controls.\nSpecific comment: In the discussion section the author gave references to other studies done elsewhere on the frequency of FF and ff genotypes, however, references on the frequency of heterozygous (Ff) is missing and should be added in the discussion section. In this study the difference was observed on this gene for the Ugandan studied population hence reference to other studies findings is important.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1890
|
https://f1000research.com/articles/5-1889/v1
|
29 Jul 16
|
{
"type": "Software Tool Article",
"title": "Automated quality control for genome wide association studies",
"authors": [
"Sally R. Ellingson",
"David W. Fardo"
],
"abstract": "This paper provides details on the necessary steps to assess and control data in genome wide association studies (GWAS) using genotype information on a large number of genetic markers for large number of individuals. Due to varied study designs and genotyping platforms between multiple sites/projects as well as potential genotyping errors, it is important to ensure high quality data. Scripts and directions are provided to facilitate others in this process.",
"keywords": [
"Genome wide association",
"genotype information",
"quality control"
],
"content": "Introduction\n\nBiases and errors can lead to erroneous associations in case-control association tests. Quality control (QC) that removes markers and individuals from a study that may introduce these biases can greatly increase the accuracy of findings. There are many examples of best practices for GWAS QC1,2. This paper describes some standard QC steps and also provides links to automated scripts to perform QC making the process easier and easily reproducible. Standard tools such as PLINK3 and SMARTPCA4,5 are called by the scripts.\n\nDue to the need for reproducibility in science, automated pipelines that can be used to repeat computational experiments and save relevant parameters is extremely important. Done step-by-step, the QC process can be quite lengthy (about 8 hours for an expert and almost certainly longer for a novice and/or someone with limited computational resources according to Anderson et al.1) and difficult to repeat exactly. Here we present scripts that perform automated GWAS QC using a parameter file that can be saved to redo the process and save human time. A log file is produced that summarizes the process to easily compare different QC parameters and their effects on the data.\n\n\nMethods\n\nQC steps implemented in this pipeline. The steps automated here mostly follow the notes on QC6 developed by MikeWeale and also calls some R7 scripts described in his notes during the QC pipeline. It is assumed that input files are already in PLINK format. Figure 1 shows a complete QC pipeline that includes combining data from multiple chromosomes and studies and two portions of the QC pipeline. There are two scripts, QC.py which takes advantage of PLINK calls and also PCA.py that does principal component analysis to investigate population stratification.\n\n1. Gender mismatches\n\nThe optional first step in the automated pipeline is a check for gender mismatch using the PLINK ‘-- check-sex’ command. This command compares the sex reported in the .fam file and the sex imputed from the X chromosome inbreeding coefficients. This step automatically removes individuals where problems are identified. The step was made optional because our dataset of interest is matched with phenotype/clinical data of higher accuracy. This step can be turned off using the parameter file as described below in the Operation section.\n\n2. Thresholds\n\nThe next steps in this pipeline include checking and applying thresholds for minor allele frequency (MAF), missingness for each individual, and missingness of markers. Minor allele frequency filtering is important because rare genotypes will not show up as often and thus will have less evidence in a GWAS and the calls will be less certain and it is also difficult to detect associations with them. Missingness can lead to false associations if it is non-random with respect to phenotypes or genotypes. Single nucleotide polymorphism (SNP) missingness is the complement to individual missingness and is correlated with SNP quality from the original genotyping assay. Missingness is investigated using PLINK ‘--missing’ and plots are generated as described by Weale6. All the plots that are generated during the process are compressed at the end in order to facilitate downloading them when the process is performed remotely on a cluster.\n\nIn order to attempt to retain the largest number of markers and individuals that pass QC there is an option to do a two-tiered missingness by individuals filtering. We noticed during testing that this could sometimes lead to final datasets with higher numbers of both. If a value is supplied to the #MIND1 parameter (described below in Operations), then this (expectedly non-stringent) threshold for PLINK ‘--mind’ is used first and the more stringent #MIND is applied in the same step as the PLINK ‘--geno’ and ‘--maf’ thresholds for missingness of markers and minor allele frequency, respectively. If a major reduction in the number of markers or individuals is found during these steps, investigation of the generated graphs can help adjust these thresholds. See notes6 for more information.\n\nSome reasoning8 suggests that a minor allele frequency threshold should be set to 10/n where n is the number of markers. The #MAF parameter can be set to ‘na’ which will use 10/n as a threshold or a threshold value can be explicitly given, such as ‘.01’.\n\n3. Heterozygosity\n\nIndividuals resulting from random mating within a population should have predictable heterozygosity (H) values. H is a measure of the number of loci in an individual that are heterozygous. Departure from expected H values can signify DNA quality issues (high H) or samples from a different population (low H). This step can be turned off by not supplying the ‘#HET’ parameter in the parameter file. As long as the parameter is listed, this step will be done. H and the inversely related F (Method-of moments F coefficient estimate) are investigated using PLINK ‘--het’. F is calculated as the ([observed homozygous count] - [expected count])/([total observations] - [expected count])) where the expected count is calculated from an imputed MAF. A histogram of F values is generated for manual investigation.\n\nIf values for ‘#FMIN’ and ‘#FMAX’ are supplied in the parameter file then samples with an F value below ‘#FMIN’ and above ‘#FMAX’ are removed. If these values are not supplied then samples above or below three standard deviations of the mean H are removed, as suggested be Anderson et al.1.\n\n4. Hardy-Weinberg equilibrium (HWE)\n\nMarkers out of HWE can indicate that there were genotyping errors. However, a strong association signal can also result in deviations from HWE. So here only variants from control samples are checked for deviations from HWE. PLINK ‘--hardy’ is used to generate HWE p-values and a Q-Q plot of the log-P-values of the markers for the controls is generated for manual investigation. A p-value threshold is supplied in the parameter file to remove markers with a p-value lower than expected.\n\n5. Cryptic relatedness\n\nCryptic relatedness (CR) is when pairs of individuals are closely related and can lead to false positive or negative correlations when subjects are treated as independent. The PLINK ‘--genome’ command can estimate relatedness, but is quite slow when there are a large number of markers in a dataset. Therefore, markers in high linkage disequilibrium (LD) are removed first to thin the data. This is done using PLINK ‘--indep-pairwise’ with parameters suggested by Weale6. This creates a pruned data set that contains markers with a minimal LD (which is caused by limited recombination occurring between two or more loci and results in a non-random association between the loci). Furthermore, only assayed markers are used in this step (i.e. not imputed markers). Using a pruned data set is advantages because CR methods work best when no LD is assumed between markers and it also reduces the input size and in turn greatly reduces the computation time.\n\nPLINK ‘--genome’ estimates relatedness of all pairs of samples and reports identify by decent (IBD, a measure of whether identical regions of two genomes were inherited from the same ancestry) in the PI_HAT (actually, proportional IBD, i.e. P(IBD=2) + 0.5*P(IBD=1)) column of the result file. A PI_HAT value close to 1 would indicate a duplicate sample. The threshold 0.1875 represents the half-way point between 2nd and 3rd degree relatives and is a common cut-off to use. Of each pair of related individuals, the one with the greater proportion of missing SNPs is dropped from the final dataset.\n\n6. Principal component analysis (PCA)\n\nGenerally, PCA transforms a data matrix (such as a GWAS n x m matrix where n in is the number of individuals and m is the number of markers and each element in the matrix represents the scaled genotype for the particular individual at that particular marker) so that the successive principal components are not correlated. The number of PCs is less than or at most equal to the original number of columns and the first PC explains the largest variance in the genotype data. Traditionally, PCA is used to (1) screen the study population for heterogeneous ethnic backgrounds and (2) to correct for potential population stratification (the difference of allele frequencies in ancestral subpopulations). It can be seen in Figure 2 where HapMap9,10 data with individuals with known ancestry are included in the PCA, when plotting the first two PCs subpopulations cluster together. HapMap is an international project that aims to identify genetic similarities and differences between populations.\n\nAs with the cryptic relatedness step, a thinned dataset created with PLINK and starting from assayed markers only is used to calculate PCs. The SMARTPCA tool is used to calculate PCs from this thinned dataset and identify outliers for removal. The PCs can then be used for further corrections in analysis models.\n\nData Formats. The input GWAS data are expected to be in PLINK bfile format. The input data will have three files associated to it with .bed, .bim, and .fam file extensions. The .bed file is a binary file that contains the genotype information for all individuals (https://www.cog-genomics.org/plink2/formats#bed). The .bim file is a mapping file giving information on each marker (https://www.cog-genomics.org/plink2/formats#bim). The .fam file gives information on each individual (https://www.cog-genomics.org/plink2/formats#fam).\n\nThis pipeline utilizes information that was not provided in the original PLINK files and therefore the phenotype is always provided in an alternate phenotype file. PLINK ‘–pheno’ is used to provide the phenotype file and PLINK ‘–pheno-name’ is used to provide the phenotype name which also corresponds to the header of the column in the phenotype file. The first two columns in the phenotype file must have the column headers ‘FID’ and ‘IID’ respectively. ‘FID’ is the family ID or ‘0’ if not used and ‘IID’ is the individual ID that corresponds to the ‘IID’ values in the .fam file (https://www.cog-genomics.org/plink2/input#pheno).\n\nSystem. The pipeline was tested on STATGEN, a Dell PowerEdge R520 server with two Intel Xeon E5-2470 CPUs (32 cores at 2.3GHz), 24TB of storage in a RAID6 array with two drive fault tolerance, and 128GB of RAM. The operating system is Ubuntu Server 14.04 LTS 64-bit edition.\n\nRequired software. The automated pipeline is written in Python and calls Rscript, PLINK, and SMARTPCA. The versions used for building and testing are the following,\n\n1. Python - Python 2.7.6\n\n2. Rscript - R scripting front-end version 3.2.2\n\n3. PLINK - PLINK v1.90b3x 64-bit\n\n4. SMARTPCA – smartpca version 13050\n\nParameter files. Example parameter files are shown in Figure 3. The QC.py and PCA.py scripts read in parameters and names them based on the word following the ‘#’ and gives that parameter the value following the white space on the same line. The line numbers in the figure are not a part of the parameter file (i.e. each line starts with ‘#’). The parameters and values are stored in a python dictionary, so order and extra parameters do not matter. However, the exact name and case of the parameters are important for the scripts to correctly function. The parameters described here are ordered by the line number given in the qc_params.txt file in Figure 3a. The parameters with the same name in the pca_params.txt file in Figure 3b have the same meaning.\n\nParameter files a) qc_params.txt b) pca_params.txt.\n\n1. WORK – path to the working directory where all generated files will be written\n\n2. RPATH – path to where Rscript is located, or name if in a known path\n\n3. PLINKPATH – path to the PLINK executable, or name if in a known path\n\n4. INPUT – input file in PLINK bfile format\n\n5. SCRIPTS – path to directory in which the helper scripts called by the pipeline live\n\n6. MAF – minor allele frequency threshold\n\n7. MIND – individual missingness threshold\n\n8. GENO – marker missingness threshold\n\n9. HWE – Hardy-Weinberg threshold\n\n10. PI_HAT – IDB threshold\n\n11. SEX – if value is equal to ‘yes’ then a PLINK sex check is performed, otherwise it is not\n\n12. ASSAYED – location of file that includes names of all assayed markers in a study to ensure that imputed SNPs are not used for some QC steps\n\n13. MIND1 – an optional, first tier (non-stringent) individual missingness threshold\n\n14. HET – if this parameter is included then the Heterozygosity step is performed (NOTE since ‘#FMIN’ and ‘#FMAX’ are not given, this parameter file will result in the removal of markers with an H value +/- 3 standard deviations)\n\n15. CLEAN – will result in the removal of intermediate files and final PLINK finals to be named ‘final’\n\n16. PHENO – the name of the phenotype that will be investigated during analysis. Some QC steps vary based on whether or not an individual is case, control, or undefined.\n\n17. PFILE – the files in which the phenotype is given formatted in the proper PLINK format with FID, IID, pheno1 column headers where FID is the family ID (commonly 0 if relations not known), the individual’s ID, and the phenotype value encoded as 1=unaffected (control) and 2=affected (case).\n\n18. SMARTPCA (line #2 in pca_params.txt and not in qc_params.txt) – link to the smartpca executable.\n\nRunning scripts and results\n\n1. QC.py\n\nOnce the parameters and values are correctly written to the parameter file, the script is executed by calling the parameter file as a command line argument as follows,\n\npython QC.py qc_params.txt\n\nThe QC steps are performed as described in the Implementation section and after each step a count is retrieved for the number of individuals or markers removed in each step. A log file is written (final_QC.log). All of the parameters from qc_params.txt are first written to the log file to ensure the process can be duplicated. Then the number of individuals/markers removed during each step are recorded along with the running total of how many individuals and markers are left in the dataset. This allows for easy comparison of the overall effects of different parameter settings. If ‘#CLEAN’ is set then the final QC dataset is in PLINK format and named ‘final.bed’, ‘final.bim’, and ‘final.fam’. There will be an archived file ‘final_graphs.tgz’ containing the files missing.png, het.png, hwe.png, and relate.png, created in the thresholds, hetereozygosity, and HWE steps above for manual inspection in case parameters need to be adjusted and QC redone.\n\n2. PCA.py\n\nOnce the parameters and values are correctly written to the parameter file, the scripts is executed by calling the parameter file as a command line argument as follows,\n\npython PCA.py pca_params.txt\n\nThe data thinning is carried out in PLINK, the input file for SMARTPCA is automatically generated, and then SMARTPCA is called to calculate the PCs. If the analysis is being done in R, then the ‘smartpca.evec’ file can be read in and merged to exclude individuals in which PCs were not calculated (i.e. outliers). PCA.py also generates a file called ‘remove.txt’ in PLINK format with an FID (all marked ‘0’) and IID column so that the outliers can be removed during PLINK analysis with the ‘–remove’ command.\n\n\nConclusions\n\nWhile the QC steps given here are not novel, this paper provides access to an automated process that both reduces human work time and chances for error and provides tools to make the computational experiment reproducible. It also gives recommended values for parameters but facilitates changing parameters and the comparison of effects.\n\n\nSoftware availability\n\nZenodo: GWAS: Automated GWAS QC, doi: 10.5281/zenodo.5822811.\n\nGitHub: https://github.com/sallyrose0425/GWAS, https://github.com/sallyrose0425/GWAS/blob/master/LICENSE",
"appendix": "Author contributions\n\n\n\nS.R.E developed automated scripts and D.W.F provided expert guidance and data. Both authors agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the National Institutes of Health (NIH) National Center for Advancing Translational Science grant KL2TR000116 and the University of Kentucky Center for Computational Sciences.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAnderson CA, Pettersson FH, Clarke GM, et al.: Data quality control in genetic case-control association studies. Nat protocols. 2010; 5(9): 1564–1573. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeale ME: Quality control for genome-wide association studies. Methods Mol Biol. Springer, 2010; 628: 341–372. PubMed Abstract | Publisher Full Text\n\nPurcell S, Neale B, Todd-Brown K, et al.: PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Human Genet. 2007; 81(3): 559–575. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatterson N, Price AL, Reich D: Population structure and eigenanalysis. PLoS Genet. 2006; 2(12): e190. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrice AL, Patterson NJ, Plenge RM, et al.: Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006; 38(8): 904–909. PubMed Abstract | Publisher Full Text\n\nWeale ME: GWAS code. Accessed: 2015-12-15. Reference Source\n\nIhaka R, Gentleman R: R: a language for data analysis and graphics. J Comput Graphical statistics. 1996; 5(3): 299–314. Publisher Full Text\n\nNeale BM, Purcell S: The positives, protocols, and perils of genome-wide association. Am J Med Genet B Neuropsychiatr Genet. 2008; 147B(7): 1288–1294. PubMed Abstract | Publisher Full Text\n\nInternational HapMap Consortium: The International HapMap Project. Nature. 2003; 426(6968): 789–796. PubMed Abstract | Publisher Full Text\n\nInternational HapMap 3 Consortium, Altshuler DM, Gibbs RA, et al.: Integrating common and rare genetic variation in diverse human populations. Nature. 2010; 467(7311): 52–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEllingson SR: GWAS: Automated GWAS QC. Zenodo. 2016. Data Source"
}
|
[
{
"id": "15320",
"date": "09 Aug 2016",
"name": "Prasoon Agarwal",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe current manuscript by Ellingson et al., is regarding the automation of the quality control (QC) of genome wide association studies (GWAS). Automation is important as it reduces human errors and increases work efficiency. In this manuscript authors have written a python script that automates the process of QC of GWAS data. The QC steps in this pipeline follow the notes developed by MikeWeale et al.\nHowever, there are some issues with the current version of the manuscript.\nThere is/are already software like QCGWAS1, which are automated for QC of GWAS. So how is the current version of the software different or better from the already existing softwares. So authors should try to summarize by comparing in terms of efficiency, accuracy in performance of their work with the already present softwares in this field.\n\nAuthors should provide a link to a test dataset that can be used to check if the required softwares are installed and running properly.\n\nThe introduction should be carefully written citing the known automated softwares for GWAS.\n\nOverall the current version of manuscript the authors have failed to bring the uniqueness of their work in the field .",
"responses": []
},
{
"id": "15797",
"date": "22 Aug 2016",
"name": "Hilary Ann Coller",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle and abstract: The title and abstract are appropriate and suitably summarize the work.\n\nArticle content: The article describes a series of scripts that the authors have developed to provide quality control measures for genomewide association studies The scripts address critical questions for genomewide association study datasets such as checking the extent of heterozygosity to determine whether it is as expected, checking whether there are markers that do not follow Hardy-Weinberg equilibrium predictions, and checking whether there are pairs of closely related individuals. The authors provide a helpful series of scripts that will allow users to perform these tasks in a reproducible and automated fashion. The scripts make it easy to determine the number of samples removed for different parameter settings. These scripts are likely to be valuable for researchers in the field.\n\nConclusions: The conclusions of the manuscript are appropriate, that use of these scripts will allow for consistent quality control of genomewide association study datasets, which will lead to better interpretation of the available data.",
"responses": []
},
{
"id": "15446",
"date": "25 Aug 2016",
"name": "Brad A. Chapman",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe a useful tool for performing automated quality control analysis of GWAS input data. They provide a set of script written in Python and R that automate running plink and smartpca to check for common issues with input data. This is a valuable contribution for documenting and automating the processes involved with checking GWAS input data.\nI have some suggestions for improving the usefulness of the tool:\nA description of the input data would help users get started with this tool. Does it take only array genotyping data or also sequencing data? Is different pre-processing or QC threshold adjustment required for the different types of data?\n\nThe software needs a minimal set of example files that users can run to confirm they have the necessary system tools installed. This will help distinguish errors in installation/running from errors in the input data. It will also help users understand the exact inputs, including plink phenotype files. It's quite hard to get started with any tool without test data. If you run into issues it is hard to distinguish differences/problems with your dataset from issues with the code.\n\nIt would be helpful for tool usage if it had a unique name. Naming the scripts generically as GWAS (or GWAS QC) makes it hard to reference or find information about the tool.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1889
|
https://f1000research.com/articles/5-1881/v1
|
28 Jul 16
|
{
"type": "Research Article",
"title": "Metagenomic analysis of an ecological wastewater treatment plant’s microbial communities and their potential to metabolize pharmaceuticals",
"authors": [
"Ian N. Balcom",
"Heather Driscoll",
"James Vincent",
"Meagan Leduc",
"Heather Driscoll",
"James Vincent",
"Meagan Leduc"
],
"abstract": "Pharmaceuticals and other micropollutants have been detected in drinking water, groundwater, surface water, and soil around the world. Even in locations where wastewater treatment is required, they can be found in drinking water wells, municipal water supplies, and agricultural soils. It is clear conventional wastewater treatment technologies are not meeting the challenge of the mounting pressures on global freshwater supplies. Cost-effective ecological wastewater treatment technologies have been developed in response. To determine whether the removal of micropollutants in ecological wastewater treatment plants (WWTPs) is promoted by the plant-microbe interactions, as has been reported for other recalcitrant xenobiotics, biofilm microbial communities growing on the surfaces of plant roots were profiled by whole metagenome sequencing and compared to the microbial communities residing in the wastewater. In this study, the concentrations of pharmaceuticals and personal care products (PPCPs) were quantified in each treatment tank of the ecological WWTP treating human wastewater at a highway rest stop and visitor center in Vermont. The concentrations of detected PPCPs were substantially greater than values reported for conventional WWTPs likely due to onsite recirculation of wastewater. The greatest reductions in PPCPs concentrations were observed in the anoxic treatment tank where Bacilli dominated the biofilm community. Benzoate degradation was the most abundant xenobiotic metabolic category identified throughout the system. Collectively, the microbial communities residing in the wastewater were taxonomically and metabolically more diverse than the immersed plant root biofilm. However, greater heterogeneity and higher relative abundances of xenobiotic metabolism genes was observed for the root biofilm.",
"keywords": [
"Wastewater treatment",
"microbial ecology",
"pharmaceuticals",
"micropollutants",
"metagenome",
"rhizoplane",
"eco-machine",
"biodegradation",
"biofilm"
],
"content": "Introduction\n\nThe treatment of human wastewater by ecological systems predates the advent of engineered wastewater treatment plants (WWTPs). While exposure to human pathogens is greatly reduced in communities where modern wastewater treatment technologies have been implemented1, widespread detection of micropollutants in the environment2 raises serious concerns about the efficacy of modern WWTPs to treat this class of contaminants. Moreover, with about two-fifths of the world’s population experience health effects due to poor sanitary conditions3. The emerging field of ecological engineering has provided a variety of viable, cost-effective wastewater treatment designs4. The organizing principle of ecological wastewater treatment is the construction of “task oriented mesocosms”5,6 of eutrophic ecosystems that, like conventional systems, primarily rely on microbial metabolic processes to achieve water quality goals. Where ecological WWTPs and conventionally engineered WWTPs differ significantly is in their reliance on ecological processes to assimilate nitrogen, phosphorous, and carbon from the wastewater into biomass. Whereas, conventional WWTPs utilize mechanically assisted, microbial processes to evolve gaseous C, N, and frequently chemical precipitation of phosphorous. A number of ecological systems have been operating around the world for decades including constructed wetlands7,8, Eco-MachinesTM4, and biofilters9,10 for residential, industrial, and municipal wastewater. These systems perform reliably based on tertiary wastewater standards4, while reducing operational costs11 and environmental and human health impacts of wastewater12. The unique potential provided by ecologically engineered waste management is direct conversion of a liability (i.e. wastewater) to an asset (sequestered carbon, biomass, products, biodiversity, etc.)13.\n\nAt the core of wastewater treatment is the biodegradation, oxidation, and reduction of organic macromolecules and inorganic chemical species primarily by resident microbial communities. While the microbial communities of conventional WWTPs have been thoroughly studied14, very little is known about microbial communities in existing ecological WWTPs despite the fact that they are central to the functions these systems provide6. The introduction of activated sludge and environmental media from diverse sources is thought to provide essential microbial functional groups4,15. It is not known whether these “seeding” events provide microbial functional groups with the capacity for biodegradation of micropollutants.\n\nThe promotion of microbial biodegradation of recalcitrant xenobiotic pollutants by plant roots has been well documented16,17. The “rhizosphere effect”18–20, driven by the release of plant metabolites from plant roots, accelerates microbial biodegradation of recalcitrant pollutants in soil and water21,22. In some cases, microbial biodegradation is promoted by a nonspecific increase in microbial metabolic activity in the area surrounding roots23, yet other studies have shown a relationship between specific plant metabolites and certain pollutant degrading organisms24. The interaction has been described as co-metabolic induction, or “co-metabolism”, where metabolism for one compound is promoted in the presence of other compounds21,22. This phenomenon has been successfully employed to accelerate the removal of a variety of recalcitrant pollutants from the soil and water including polychlorinated biphenols (PCBs)19,20,24, polycyclic aromatic hydrocarbons22, and chlorinated solvents such as trichloroethylene25. However, little research has been done on whether co-metabolism is occurring in ecological WWTPs as a result of plant-microbe feedback processes. Using whole metagenome sequencing (WMS) we have examined whether the microbial populations residing on the plant roots immersed in wastewater of an ecological WWTP showed evidence of the capacity for micropollutantant biodegradation. These populations were compared to microbial communities free-floating in the wastewater, enrichment cultures growing on individual pharmaceutical compound carbon sources, as well as PPCP concentrations throughout the treatment system to determine whether plant-microbe feedback processes are supporting PPCP biodegradation in ecological WWTPs.\n\n\nMaterials and methods\n\nCarbamazepine (5H-Dibenz[b,f]azepine-5-carboxamide), sulfamethoxazole (4-Amino-N-(5-methyl-3-isoxazolyl)benzenesulfonamide), and trimethoprim (2,4-Diamino-5-(3,4,5-trimethoxybenzyl)pyrimidine) were purchased from Sigma Aldrich.\n\nAn ecological WWTP (Eco-MachineTM) in Sharon, Vermont at the Vietnam Memorial rest area on northbound Interstate 89 (43.727896, -72.425564) was sampled on June 30 and July 1, 2013. Wastewater from the toilets, urinals, and sinks, is collected in a holding tank and then treated in a series of tanks (Figure 1). These consist of an anoxic tank (ANOX), a closed tank (CLO) and planted aerobic tanks (HR1, HR2 and HR3). These are followed by a clarifier and final treatment by a sand filter (SAND). The treated water (effluent, hereafter) is disinfected with the addition of sodium hypochlorite and dyed blue prior to returning to the toilets and urinals for reuse. To accommodate the approximately 48 hour residence time of the wastewater in the system [personal communication-Phil Gates, Simon Management Services], samples of aqueous phase and the immersed biofilm were collected from the first three tanks (ANOX, CLO, HR1) on June 30, 2013 and the latter three tanks (HR2, HR3, and SAND) on July 1, 2013. Plant root biofilm samples consisted of multiple roots from each individual tank composited into one sample. Influent wastewater samples (INF) were collected from the holding tank on June 30th.\n\nDuplicate 1 L aqueous phase samples were collected from each of the treatment tanks as well as the system INF and effluent (EF) for quantification of PPCPs by EPA method 169426,27 at a commercial analytical lab (TestAmerica, Sacramento, CA) using the Waters Acquity UPLC System and Waters Micromass Quattro Premier XE Mass Spectrometer.\n\nEnrichment cultures with the pharmaceutical compounds carbamazepine, trimethoprim, and sulfamethoxazole (0.1M) serving as individual carbon sources were initiated using wastewater effluent inoculum in 100 mL carbon-free mineral salts medium (10 mM KH2PO4, 3 mM NaH2PO4, 1 mM MgSO4, 1mM NH4SO4 and trace minerals28). Carbamazepine was delivered with minimal amounts of methanol added to the flask immediately after autoclaving and was allowed to evaporate leaving small suspended crystals as the sole carbon source. Starting with 1 mL of the WWTP sample, enrichment cultures were maintained at room temperature in a rotary shaker (100 rpm) for approximately 90 days. Five replicate cultures were initiated for each individual pharmaceutical carbon source.\n\nAt the third serial enrichment samples from the carbamazepine cultures (C3A, C3B, and C3D), trimethoprim cultures (T3B, T3C, and T3D) and sulfamethoxazole cultures (S3B, and S3D) were selected based on visual verification of microbial growth in the flasks. These eight samples were used for all further analyses.\n\nTotal genomic DNA was extracted and combined from duplicates for all samples using the following methods: water, biofilm, and enrichment culture samples were centrifuged at > 8,000 g for 1 min. Excess liquid was removed and pellets containing microbial samples were homogenized. Homogenization was performed using ~300 mg of a 50/50 mix of 1 mm and 100 μm AlO3 abrasive and 1 1/4 mm ceramic ball (Matrix F equivalent-MP Biomedical) and FastPrep-24 (MP Biomedicals, Santa Ana, CA) for 20 sec. at 6.5 R/S. 10 μL of 10 μg/μL lysozyme (Sigma), 4 μL of 400 U/μL Achromopeptidase (Sigma), 2 μL Mutanolysis (5U/μL) prepared in 10 mM TRIS buffer were added to each sample, which were briefly vortexed and incubated overnight at 37°C. The samples were then extracted using the standard method outlined by the E.Z.N.A.® Mollusc DNA isolation kit (Omega-Biotek, Inc, Norcross, GA), and the resulting DNA was quantified and its quality was assessed using the Nanodrop spectrophotometer (Thermo Scientific, Madison, WI), and Qubit Spectrofluorometer (Life Technologies, Carlsbad, CA) according to manufacturer’s instructions. After duplicate samples were combined, the resulting DNA concentrations were between 1.1 ng/μL for the wastewater samples obtained from tanks HR1 and HR2 and 17.9 ng/μL for the biofilm sample collected from the anoxic tank. Fragmentation of 10–100 ng of the resulting DNA was performed using a Covaris S2 AFA sonicator (Covaris Corp., Woburn, MA) equipped with MicroVails (http://covarisinc.com/products/afa-tubes-and-vials/microtube-15/) to yield a size range of 200–500 bp as confirmed through a high sensitivity microfluidic DNA chip on the Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA) according to the manufacturer's instructions. The Agilent 2100 Bioanalyzer is an automated microfluidic-chip that is widely used to assess the DNA size fragment distribution and quantification in next-generation sequencing.\n\nLibrary preparation was performed using 45 ng of DNA (except samples HR1_W and HR2_W, which produced a total of 33 ng of DNA) in accordance with the Illumina® TruSeq DNA Sample Prep LT version 2 SOP (Part # 15026486 Rev. C, July, 2012) with the indicated reagents (DNA kit #FC-121-2001). According to manufacturer’s instructions, each sample was subjected to end repair, adenylation, and ligation of Illumina adaptors for indexing purposes. PCR amplification was performed using Illumina reagents (Part#15012995) followed by quantification using the Qubit spectrofluoromter and qPCR quantitation kit (KAPA Biosciences kit # 4824). Library quality and insert size distribution was assessed using the Agilent Bioanalyzer 2100.\n\nCluster generation and paired-end sequencing were performed at the Delaware Biotechnology Institute (DBI), University of Delaware, using an eight lane high-capacity v3 flowcell on the Illumina cBOT and HiSeq 2000 sequencer (Illumina, San Diego, CA), respectively. The WWTP samples (twelve) and the enrichment culture samples (eight) were multiplexed and run on two lanes. DBI delivered 20 FASTQ files with raw sequence data.\n\nRaw sequences were checked for quality with FastQC v0.10.1 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Trimmomatic v0.3029 was used to remove adapters and filter low-quality base calls/reads. Leading and trailing bases below quality 20 and reads less than 40 bases in length were removed. Additionally, reads were scanned using a 5-base wide sliding window and cut when the average quality per base dropped below 20. PhiX Control v3 from Illumina was used as a low-concentration spike-in during sequencing at DBI. Quality-trimmed FASTQ files were aligned to the PhiX genome (NCBI RefSeq NC_001422.1) using Bowtie2 2.2.330 and all aligned reads were removed. Quality-trimmed and filtered reads were verified with FastQC prior to taxonomic and functional characterization.\n\nAll twenty FASTQ files and associated metadata are available through NCBI BioProject ID PRJNA286671 (http://www.ncbi.nlm.nih.gov/bioproject/286671).\n\nTranslated trimmed reads served as input for a protein-level homology search against NCBI-NR, (ftp://ftp.ncbi.nlm.nih.gov/blast/db/FASTA/nr.gz, downloaded May 26, 2015) a comprehensive non-redundant protein database, using the BLAST-like tools RAPSearch2 v2.1631 for WWTP samples and DIAMOND v.0.7.932 for enrichment culture samples. DIAMOND was used instead of RAPSearch2 for analysis of enrichment culture samples because it was designed to easily integrate with MEtaGenome ANalyzer (MEGAN). It implements an algorithm that is similar to, but faster than, RAPSearch2, it was newly available when the enrichment cultures’ sequence data was ready for analysis, and control sample testing showed nearly identical taxonomic profiles from DIAMOND as those generated with RAPSearch2 searches.\n\nThe similarity search results for each sample set, which include all reads with alignments to the NR protein database and their GI accession numbers (maximum 25 alignments per read) were imported separately into MEGAN v5.7.1033 (http://ab.inf.uni-tuebingen.de/software/megan5/). MEGAN parsed the RAPSearch2 (WWTP) and DIAMOND (enrichment culture) results using the lowest common ancestor (LCA) algorithm34 and NCBI taxonomy (ftp://ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdmp.zip downloaded March 26, 2015) (lowest common ancestor parameters: maxMatches=100 minScore=50.0 maxExpected=1.0 topPercent=10.0 minSupportPercent=1.0 minSupport=50 minComplexity=0.44). Reads that passed this filter and that were unambiguously assigned to a NCBI taxon by LCA were retained in each sample’s MEGAN results file. A combined MEGAN file was generated for WWTP samples, as well as for enrichment culture samples, with read counts normalized to the sample with the fewest input reads in each set.\n\nTwo positive controls were used to validate our bioinformatics pipeline and to establish a minimum support threshold (or false-positive cut-off) for taxonomic profiling. One control dataset is comprised of single-end Illumina reads from a synthetic microbial sample prepared by CosmosID. This constructed freshwater sample simulates organisms found in the Delaware River and is described here: http://www.cosmosid.net/constructed-freshwater. The second control dataset is single-end Illumina reads from the Human Microbiome Project (HMP) mock community even sample (http://www.ncbi.nlm.nih.gov/sra/SRX055380). Reads from both positive control samples can be downloaded from BaseSpace: https://basespace.illumina.com/projects/20039022/samples.\n\nUsing MEGAN, reads were annotated based on the KEGG (Kyoto Encyclopedia of Genes and Genomes) functional classification of enzymes and pathways36. Using auxiliary index files obtained from the MEGAN website (gi2kegg.map.gz, built Dec 1, 2010), GI accession numbers were mapped to KEGG functional groups based on the highest MinScore match (minimum MinScore = 50). Reads may be assigned to more than one functional group per classification system, as each KEGG group may appear in several functional categories.\n\nThe relatedness of the microbial communities located in the different tanks and phases of the WWTP was assessed through pairwise similarity scores computed in MEGAN using a normalized Goodall’s probabilistic similarity index38 for both phylogenetic and metabolic profiles for each sample. Graphical representations of the distance matrices were generated in MEGAN as un-rooted phylogenetic neighbor networks39. A Venn diagram was produced (Partek® Genomics Suite® software, version 6.6 build 6.15.1016 Copyright; 2014, Partek Inc., St. Louis, MO, USA) to illustrate taxa common to the different sample datasets.\n\nStatistical Analysis of Metagenomic Profiles (STAMP) software v2.1.340 was used to test statistical significance of differentially abundant taxonomic groups and functional categories for 1) WWTP sample groups (aqueous and biofilm phases) and 2) enrichment culture sample groups (carbamazepine (C), sulfamethoxazole (S), and trimethoprim (T)). LCA taxonomic profiles and KEGG including abundances, were imported to STAMP for each sample set. Two-sided Welch's t-test was used to compare aqueous and biofilm phases with a confidence interval of the effect size and multiple test correction using the Benjamini-Hochberg FDR method. One-way ANOVA was used to compare enrichment culture groups with an effect size (Eta-squared) and multiple test correction using the Benjamini-Hochberg FDR method. Tukey-Kramer post-hoc test (0.95) was used to determine which means were significantly different when an ANOVA produced a significant p-value.\n\nTo visualize the distribution of microbial taxa in WWTP samples the Circos software package v0.6941 was used to depict the location and relative abundances of microbial taxa at the class rank identified in MEGAN using the LCA algorithm.\n\n\nResults\n\nThere were 11,568 visitors during the week in which sampling was conducted (June 25–July 1, 2013). Each visitor used an average of 2.27 liters of water contributing 26,452 liters of water to the wastewater treatment system42. The wastewater used to isolate microbial DNA samples contained detectable concentrations of caffeine, carbamazepine, DEET, gemfibrozil, ibuprofen, naproxen, sulfamethoxazole, thiabendazole, and trimethoprim. Of the compounds detected in influent water (from facility toilets, urinals, and sink drains), the cholesterol medication gemfibrozil was detected at the highest concentration (1.5 × 105 ng L-1), followed by caffeine, ibuprofen, and naproxen (9.5, 6.6, and 5.5 × 104 ng L-1, respectively) (Table 1). The concentrations of PPCPs in the wastewater samples generally decreased the further through the treatment process (Figure 1) the sample was obtained. However, gemfibrozil, caffeine, and ibuprofen were detected at higher concentrations in the sand filter or effluent water samples than the preceding tank. The concentrations of carbamazepine, DEET, and trimethoprim did not change substantially over the entire treatment process.\n\nConcentrations (ng L-1) of detected pharmaceuticals and personal care products in the wastewater sampled from each major treatment tank of the WWTP. Abbreviations: INF- influent, ANOX- anoxic closed tank, CLO-closed aerobic tank, HR1, HR2 & HR3- planted aerobic tanks, SF- sand filter, EF- fffluent, Caff- caffeine, Carb- carbamazepine, DEET - N,N-Diethyl-3-methylbenzamide, Gemf- gemfibrozil, Ibup- ibuprofen, Napr- naproxen, Sulf- sulfamethoxazole, Thia- thiabendazole, Trim- trimethoprim, ND-not detected above method reporting limit.\n\nAbbreviations: Hold – holding tank, ANOX- anoxic tank, CLO- closed tank, HR1, HR2 and HR3- planted aerobic tanks, CLF- clarifier, SF sand filter.\n\nWhole metagenome shotgun sequencing of 12 WWTP samples generated more than 388 million paired-end reads, 101 bp in length, with an average depth of 32.4 million reads per sample (range: 2.35–53.3 million) (Supplementary Material ST13). Eighty-eight percent of raw reads (343,355,560) were retained after quality-trimming and were aligned to the NCBI-NR protein database. Of the 175,238,945 reads with at least one hit to NR proteins, approximately 84% were assigned taxonomy by the lowest common ancestor (LCA) algorithm in MEGAN. Over half of quality-trimmed reads in our samples (51%) had no protein hits in NCBI-NR and 19.3% of reads with protein hits could not be classified by the LCA algorithm. As a result, the latter were designated “Not Assigned” reads in MEGAN.\n\nWhole metagenome shotgun sequencing of the 8 enrichment culture samples generated more than 177 million paired-end reads, 101 bp in length, with an average depth of 22 million reads per sample (range: 17–29 million) (Supplementary Material ST14). Eighty-nine percent of raw reads (157,598,266) were retained after quality-trimming and were aligned to the NCBI-NR protein database. Of the 96,824,600 reads with at least one hit to NR proteins, over 99% were assigned taxonomy by the LCA algorithm in MEGAN. Nearly 39% of quality-trimmed reads in our samples (60,773,666) had no protein hits in NCBI-NR and 0.4% of reads with protein hits could not be classified by the LCA algorithm and were designated “Not Assigned“ reads in MEGAN. The minimum-support percent threshold in MEGAN for both WWTP and enrichment culture analyses was set to 1.0% based on our bioinformatics workflow results from the HMP and Delaware River control samples.\n\nThe LCA algorithm provided a microbial taxonomic profile of the 12 WWTP samples and the 8 enrichment culture samples (Figure 2 and Figure 3, respectively). The read counts were normalized to the sample with the least number of total reads to allow relative abundance to be depicted and shown as bar-charts at the leaves of each phylogram. The terminal taxa from various ranks identified by the LCA algorithm in the 12 WWTP samples ranged from a low of 8 taxa (HR1_B sample) to 17 taxa (ANOX_W sample) (Figure 2; Supplementary Material ST2). Members of Mycobacterium, Pseudomonas, and Verrucomicrobia were identified in all of the aqueous phase (_W) samples as well as the biofilm in the anoxic tank (ANOX_B) and the sand filter (SAND_B). Additionally, the families Rhodobacteraceae, Burkholderiaceae, Comamonadaceae and Xanthomonadaceae were identified in all of the aqueous phase samples as well as in at least one, but not all immersed biofilm samples. The family Xanthomonadaceae was identified in 10 of the 12 WWTP samples, followed by Rhodobacteraceae (9), Comamonadaceae and Rhizobiaceae (7), and Burkholderiaceae (6).\n\nPhylogram depicting the lowest common ancestor taxonomic composition of the ecological wastewater treatment plant. Bar chart for each taxon (depicted in the order shown in the legend) indicate the number of reads (normalized) associated with each taxonomic classification, shown here in square-root scale to highlight differences. Wastewater treatment plant sample locations with _W and _B indicate aqueous and immersed biofilm samples, respectively. Abbreviations: ANOX- anoxic tank, CLO- closed aerobic tank, HR1, HR2 and HR3- planted aerobic tanks, and SAND - sand filter.\n\nPhylogram depicting the lowest common ancestor taxonomic composition of the sole pharmaceutical compound carbon source enrichment cultures. Bar charts for each taxon (depicted in the order shown in the legend) and indicate the number of reads (normalized) associated with each taxonomic classification, shown here in square-root scale to highlight differences. Samples starting with C, T, and S indicate sequences obtained from cultures grown on carbamazepine, trimethoprim, and sulfamethoxazole carbon sources, respectively.\n\nAccording to the LCA algorithm taxonomic assignments by MEGAN, the enrichment cultures originating from the wastewater effluent inoculant produced mixed cultures ranging from 2 (T3B and T3C) to 8 (S3D) taxa identified in each culture (Figure 3; Supplementary Material ST4). Bacteria were identified in all enrichment cultures and ascomycete fungi were in all carbamazepine cultures and one trimethoprim culture (T3D). The family Nocardioidaceae in the Propionibacterineae was identified in all but two of the cultures (C3D and S3B). Proteobacteria were identified in all but one of the carbamazepine cultures (C3B). Alphaproteobacteria was identified in two carbamazepine cultures (C3A and C3D), all sulfamethoxazole cultures as well as one of the trimethoprim cultures (T3D). Gammaproteobacteria was identified in one sulfamethoxazole culture (S3B). Numerous taxa were identified at the species level by LCA in MEGAN, including Bacillus cereus, B. lichenfomis, B. subtilis, Clostridium perfringens, Hyphomicrobium denitrificans, H. nitrativorans, H. zavarzinii, Sphingobium sp. SYK-6, Trichoderma atroviride, T. virens, and Meyerozyma guilliermondii.\n\nThe carbamazepine cultures (C3A, C3B, and C3D) all contained members of Actinomycetales, Ascomycete fungi, while Firmicutes and Alphaproteobacteria are represented in two of three replicates (C3B, C3D and C3A, C3D, respectively). Similarly, the trimethoprim enrichment cultures (T3B, T3C and T3D) contained Actinomycetes and Proteobacteria with Alphaproteobacteria and ascomycete budding yeast fungus Aspergillus in culture T3D only. The sulfamethoxazole enrichment cultures contained highest taxonomic richness with eight taxa identified by LCA in MEGAN in one of the replicates (S3D) and seven taxa in the other (S3B). These taxa collectively included members of the Actinomycetales (Leucobacter), Rhizobiales (Bradyrhizobiaceae, Hyphomicrobium, Mesorhizobium, Rhizobium/Agrobacterium group), and Gammaproteobacteria.\n\nA graphical representation of the pairwise distance matrix39 generated using normalized Goodall’s similarity index38 of the 12 WWTP samples is shown as an unrooted phylogenetic neighbor network in Figure 4. The LCA taxonomic assignments of the microbial populations residing in the aqueous phase samples cluster near one another in the neighbor network, while the immersed biofilm samples showed greater dissimilarity.\n\nNeighbor-net depicting the taxonomic pairwise similarity (Normalized Goodall) of the lowest common ancestor of translated sequences obtained from the six major tanks of the WWTP and the two major phases (aqueous _W and immersed biofilm _B) of the ecological wastewater treatment plant. Abbreviations: ANOX - anoxic tank, CLO - closed aerobic tank, HR1, HR2 and HR3 - planted aerobic tanks, and SAND - sand filter.\n\nIn STAMP, the aqueous and biofilm samples were compared at different taxonomic ranks to highlight differences between the WWTP sample sets. Figure 5 depicts the percent relative abundances of microbial classes identified in the 12 WWTP samples. In general, Actinobacteria, Alphaproteobacteria, Betaproteobacteria, and Gammaproteobacteria had higher relative abundances in the aqueous samples, whereas Bacilli were more abundant in the biofilm samples. The relative abundances of Bacilli in the biofilm samples were highest in the first three tanks (ANOX, CLO, HR1) and sharply reduced thereafter (Figure 5a, Figure 9). The relative abundances of Deltaproteobacteria, Betaproteobacteria, Gammaproteobacteria, and Bacilli were significantly different between the two physical phases (aqueous and biofilm) sample sets (Figure 5b).\n\nA. Heat map depicting the relative abundance of microbial taxonomic classes identified in the different aqueous and biofilm phases in each treatment tank of the ecological wastewater treatment plant. B. Differences in mean proportions of major microbial taxonomic classes identified collectively in the aqueous and biofilm phases of the ecological wastewater treatment plant. Statistical Analysis of Metagenomic Profiles (STAMP) software v2.1.3. Abbreviations: W – aqueous, B – immersed biofilm, ANOX – anoxic tank, CLO – closed aerobic tank, HR1, HR2 and HR3 – planted aerobic tanks, and SAND – sand filter.\n\nTo assess similarities in taxonomic composition between the pharmaceutical compound enrichment cultures and the WWTP samples, we examined the co-occurrence of LCA taxa among datasets (Figure 6). The taxonomic similarities of the five major sample types were examined by combining all aqueous samples, all immersed biofilm samples, and the replicates of each pharmaceutical carbon source enrichment cultures into five datasets. As is typical of enrichment cultures originating from complex environmental samples, taxonomic similarities were limited to a few shared taxa. When combined, twelve microbial taxa were identified in the carbamazepine enrichment cultures. However, only three taxa were identified in at least two out of three replicate carbamazepine cultures. Of the twelve taxa identified in the combined carbamazepine cultures, only one each was identified in the biofilm and aqueous samples (Bacillus cereus, Mycobacterium spp., respectively) obtained from the wastewater treatment plant. Bacillus cereus was identified in the biofilm samples obtained from the first three tanks of the wastewater treatment system (ANOX_B, CLO_B, and HR1_B). It has been associated with human feces43. One taxon (Nocardioidaceae bacterium Broad-1) was found in six out of eight total enrichment culture samples, including all three trimethoprim samples. Additionally, Proteobacteria, Aspergillus spp. and Methylobacterium spp. were unique to trimethoprim enrichment cultures. Nocardioidaceae bacterium Broad-1, found as a byproduct during the genome assembly of the fungus Coccidioides (NCBI BioProject accession number PRJNA48513), is of unknown origin. Of the ten taxa identified in the combined sulfamethoxazole cultures, 50% were in both sulfamethoxazole samples, three were identified in the biofilm samples (Rhizobiaceae (HR2_B, HR3_B and Sand_B), Bradyrhizobiaceae (Sand_B), and Mesorhizobium spp. (HR2_B)), and one was found in the all of the aqueous samples (Rhizobiaceae) obtained from the wastewater treatment plant. Seven out of 40 WWTP taxa were shared among the aqueous and biofilm samples.\n\nVenn diagram illustrating sequences with taxonomic identities retrieved from the aqueous and biofilm samples obtained from the ecological wastewater treatment plant and the carbamazepine, sulfamethoxazole and trimethoprim aqueous enrichment cultures. Numbers in parentheses after sample set names indicate taxa identified to species level in each metagenome sequence dataset. Numbers within the diagram indicate taxa common to the individual and overlapping datasets. (Partek® Genomics Suite® software, version 6.6 build 6.15.1016 Copyright 2014; Partek Inc., St. Louis, MO, USA)\n\nTo assess the influence of the carbon sources on enrichment culture taxonomic composition, LCA-based taxonomic profiles of the enrichment culture replicates were compared in STAMP. The most numerous significant differences with strong effect sizes were found at the family level (see supplementary material ST35). The relative abundance of reads from Bradyrhizobiaceae, Hyphomicrobiaceae, Microbacteriaceae, and Rhizobiaceae were significantly greater in sulfamethoxazole cultures than in trimethoprim or carbamazepine cultures (with FDR corrected p-values of 2.98 × 10-8, 0.0274, 0.0199, 0.381, respectively).\n\nTo investigate the role of the WWTP’s microbial communities in the metabolism of PPCPs, KEGG was queried using MEGAN for sequences identified as involved in xenobiotic metabolism. Figure 7 depicts the percent relative abundances of sequences identified in the metagenomes obtained from each sample as involved in xenobiotic biodegradation and metabolism KEGG pathways for WWTP samples. The most abundant xenobiotic biodegradation and metabolism subcategory in all cultures and WWTP samples, with a few exceptions (HR1_B, C3B and C3D), was benzoate degradation comprising 14.5%–17.7% of reads in this subcategory for WWTP samples and 11.1%–23.6% in the enrichment culture samples. The biofilm sample obtained from the sand filter (Sand_B) at the end of the WWTP treatment train produced the greatest number of sequences (18,766) aligning to the “xenobiotic biodegradation and metabolism” category in the KEGG database followed by the aqueous samples of the sand filter (SAND_W) (18,318), and the aqueous phase of the third planted aerobic tank (HR3_W) (18,002). The sample with the greatest number of reads associated with KEGG pathway “drug metabolism - cytochrome P450” was the aqueous phase of the third planted aerobic tank (HR3_W) (1,384) followed by aqueous samples of the sand filter (SAND_W) (1,372) and the immersed biofilm sample obtained from the first planted aerobic tank (HR1_W) (1,330). The sample with the greatest number of reads associated with the KEGG category “drug metabolism – other enzymes” was the biofilm in the second planted aerobic tank (HR2_B) (2,181) followed by aqueous phase of the anoxic tank (ANOX_W) (1,904), and the biofilm sampled from the sand filter (SAND_B) (1,827).\n\nHeat map indicating the relative number of sequence reads associated with the KEGG xenobiotic metabolism categories (on right) identified in each sample location and phase (on bottom). Statistical Analysis of Metagenomic Profiles (STAMP) software v2.1.3. Abbreviations: W – aqueous, B – immersed biofilm, ANOX – anoxic tank, CLO – closed aerobic tank, HR1, HR2 and HR3 – planted aerobic tanks, and SAND – sand filter.\n\nThe percent relative abundances of xenobiotic metabolism-associated sequences identified in enrichment cultures’ metagenomes are illustrated as a heat map in Figure 8 (see Supplementary materials ST6). Reads associated with benzoate degradation were most abundant for trimethoprim cultures, whereas chloroalkane and chloroalkene degradation reads were most abundant for sulfamethoxazole cultures. The most abundant xenobiotic metabolism category for the carbamazepine cultures varied from culture to culture with aminobenzoate degradation, benzoate degradation, and chloroalkane/chloroalkene degradation categories for each of the three replicates. The samples C3A and T3D produced the greatest number of reads associated collectively with xenobiotic metabolism genes at 232, 201 and 218, 591, respectively. Of these, C3A and T3D metagenomes contained 10,151 and 11,889 reads associated with “drug metabolism – other enzymes” in KEGG, respectively. In this category, the sulfamethoxazole culture (S3B and S3D) had the most assigned reads at 16,102 and 13,730, respectively. For the category “Drug metabolism – cytochrome P450” the carbamazepine culture sample C3A contained the greatest relative number of sequences (16,199) followed by the sulfamethoxazole culture S3D (11,847).\n\nHeat map indicating the relative number of sequence reads associated with the KEGG xenobiotic metabolism categories (on right) identified in each pharmaceutical compound carbon source enrichment culture sample (on bottom). Samples starting with C, T, and S indicate metagenomes isolated from cultures grown on carbamazepine, trimethoprim, and sulfamethoxazole carbon sources, respectively. Statistical Analysis of Metagenomic Profiles (STAMP) software v2.1.3.\n\n\nDiscussion\n\nTo our knowledge, this is the first study to use WMS to characterize the microbial communities of an ecological WWTP. While a taxonomic description of the microbial communities is provided, here we have focused on microbial metabolism of PPCPs in the wastewater. Using WMS we have identified representation and relative abundance of microorganisms in the six major treatment tanks. To examine the role plants have on the structure of the microbial communities, we compared the communities found in the wastewater and those attached to plant roots immersed in the wastewater. Microbial metabolic pathways for most emerging pollutants, including micropollutants such as PPCPs, have not been characterized. Therefore, we have focused on the xenobiotic metabolic capacity as represented by the location and abundance of known genes represented in the KEGG database.\n\nThe concentrations of PPCPs in the samples obtained from the Sharon, VT WWTP indicate that some of the PPCPs are being effectively removed by the system (naproxen and thiabendazole), while others are accumulating in the recirculating system (caffeine, carbamazepine, DEET, trimethoprim, and sulfamethoxazole) (Table 1). The increasing concentrations of caffeine, gemfribozil and ibuprofen in or after the sand filter suggest partitioning to the aqueous phase from attenuated organic matter may be occurring in the sand filter. However, based on these data, biotic (biodegradation) and abiotic (partitioning to the primary sewage sludge) processes occurring in the influent holding tank and first treatment tank (ANOX) account for significant reductions in concentrations for the PPCPs gemfibrozil, naproxen, thiabendazole, and to a lesser degree caffeine, ibuprofen, sulfamethoxazole, whereas carbamazepine, DEET, and trimethoprim were not removed.\n\nPartitioning to the solid phase (sewage sludge) is an important aqueous phase removal processes that is driven by a compound’s hydrophobicity44,45. For certain chlorinated organic compounds, the octanol-water partitioning coefficient (KOW) correlates positively with sorption to biosolids when log KOW values range from 1.26 to 5.4846. Carbamazepine (log KOW 2.3–2.5), DEET (log KOW 2.18–2.44), and trimethoprim (log KOW 0.9–1.4) show limited removal in the Sharon, VT WWTP (Table 1). These findings are consistent with similar studies on partitioning of PPCP in conventional WWTP47–49. Partitioning rates from aqueous influent to biosolids (sewage sludge) is variable with log Kd (solid-water distribution) values ranging from <0.7 to 4.250 depending on the compound. The removal of the parasiticide, fungicide thiabendazole (log KOW 2.47) from the wastewater between the influent sample location and the anoxic treatment tank was likely due to strong partitioning to the primary sludge in the holding tank. With this exception, all other compounds were detected in more than one tank of the system. The aqueous concentrations of the cholesterol-lowering drug gemfribozil (log KOW 4.77) also indicated strong partitioning to the primary sludge showing reduction in concentration by three orders of magnitude between the influent and the anoxic treatment tank.\n\nPharmaceuticals and PPCPs in wastewater can undergo a number of processes that contribute to their complete or partial aqueous phase removal in wastewater treatment systems45. These include chemical and or physical processes such as sorption to organic matter46,48,51, photolysis52,53, volatilization50, and biological transformation54. Biological transformation is unique among these as it provides WWTP operators the potential to increase the removal of PPCPs from wastewater while partially or completely mineralizing the compounds thereby eliminating any risks associated with their release to the environment. In contrast, sorption to biomass (primarily sewage sludge) results in decreased mineralization55, which when applied to land (dominant disposal method of the processed sewage sludge) is likely a significant source of PPCPs in the environment56. It is unclear whether thermal treatment and dewatering, as is commonly done to biosolids prior to land application, alters the mass of PPCPs in this media.\n\nThe influent concentrations of the PPCPs were in many cases an order of magnitude or greater than the concentrations reported in conventional WWTP49,57. The Sharon, VT ecological WWTP recirculates the effluent onsite as flush water (sterilized and dyed blue prior to being used as toilet and urinal flush-water). The recirculation and reuse of the effluent is likely to result in an additive or concentrating effect for compounds with low removal and/or partitioning rates. Additionally, the concentration of PPCPs in municipal wastewater is likely diluted by the mixing of non-human wastewater such as wash water, storm water (in combined sewer systems), industrial process water and a variety of other sources. While the higher concentration of PPCPs in this recirculating ecological WWTP may present elevated exposure risks to operators and the environment if materials are discharged, conventional WWTP, which do not recirculate wastewater, are likely to discharge greater mass of PPCPs per liter wastewater treated. Additionally, retaining the PPCPs in the WWTP through recirculation is preferential to releasing them into receiving water bodies.\n\nThe results reported here are initial findings on the removal of PPCPs from the wastewater processed by this system as the species and concentrations of detected PPCPs are likely to change with time, fluctuating with the changing population of visitors. Significant variability in PPCPs species and mass loading into the system is likely responsible for the non-detects (Table 1 ND’s) of individual compounds detected later in the treatment train. As we staggered our sampling of the initial three and latter three treatment tanks by 48 hours to accommodate the residence time of the wastewater in the treatment system, it is reasonable to assume that the concentrations quantified here represent the flux of PPCPs through the WWTP. Therefore, changes in the concentration of individual PPCPs throughout the system are the result of biotic and abiotic aqueous phase removal processes. The trends observed here could be influenced by fluctuations in PPCPs inputs. These fluctuations are likely responsible for the non-detects observed in the holding tank (INF) for carbamazepine, DEET and trimethoprim, while these compounds were detected in the next treatment tank (ANOX).\n\nThe immersed biofilm and aqueous phase microbial communities exhibited two distinct taxonomic structures. According to the LCA algorithm, this difference was most evident at the class level (Figure 5a). The relative abundances of Deltaproteobacteria, Betaproteobacteria, and Gammaproteobacteria were significantly higher in the microbial metagenomes of the aqueous phase, while Bacilli were observed in greater abundance in the immersed biofilm microbial communities (Figure 5b). The relative abundance of Bacilli in the immersed biofilm communities was highly variable, with their dominance diminishing significantly after the first three tanks. Figure 9 illustrates the microbial taxonomic spatial variability of the ecological WWTP. The sample order in the diagram reflects wastewater movement through the treatment system starting with the anoxic tank (ANOX) moving clockwise to the sand filter (SAND). The width of each ribbon extending from the taxon to the sample represents relative abundance based on sequence counts. Bacilli dominated the immersed biofilm in the first three tanks (ANOX, CLO, and HR1), with representatives of other taxa present to a much lesser degree. The immersed biofilm samples also show greater microbial taxonomic richness in the latter phases of the treatment system (HR2, HR3, and SAND). This pattern is likely due to retention of fecal taxa by the earlier tanks as Bacilli are known to be abundant in human feces43. Bacteroidia, which is also abundant in human feces58, was not detected colonizing the immersed plant root surfaces. Only a small population was identified in the wastewater obtained from the anoxic tank.\n\nRelative abundances of microbial class-level taxa identified in the metagenomes isolated from the biofilm and wastewater phases of each treatment tank of the ecological wastewater treatment system. The relative abundance of each class in each sample is represented in the width of each ribbon. The clockwise order of the samples is represents the order of the wastewater treatment process. Circos software v0.69. Abbreviations: ANOX- anoxic tank, CLO- closed aerobic tank, HR1, HR2 and HR3- planted aerobic tanks, and SAND - Sand Filter.\n\nMembers of the Firmicutes (Bacilli) and Bacterioidia, which are common in human feces58, were identified in the ecological WWTP microbial communities. However, these were the only organisms identified in the samples that are associated with human feces. Of the organisms used in water quality criteria to indicate contamination by feces (others include Clostridium perfringens, enterococci, Escherichia coli, and fecal coliforms), Clostridium perfringens was detected in only one sample, HR1_B (Figure 2). This would indicate that the ecological WWTP is performing well with regard to attenuating fecal coliforms. Bacilli were dominant in the samples of immersed biofilm collected from the first three tanks in contrast to the taxonomic composition of the latter three treatment tanks (Figure 2, Figure 9). This could indicate colonization of the surfaces in these initial treatment tanks by organisms of fecal origin. The wastewater samples from these tanks as well as biofilm samples collected from tanks later in the treatment train did not show this pattern (Figure 9).\n\nThe microbial taxonomic composition of the immersed biofilm in the downstream tanks increased in richness with Actinobacteria, Alphaproteobacteria, Bacilli, Gammaproteobacteria, Sphingobacteria and to a lesser degree Betaproteobacteria, and Planctomycetia identified by LCA. This is in contrast to the aqueous phase microbial community, which showed taxonomically rich populations throughout the system with members of the Actinobacteria, Betaproteobacteria, Alphaproteobacteria, and Gammaproteobacteria in greatest abundance (Figure 9).\n\nThe taxonomic composition of the microbial communities forming biofilm on the plant root surfaces immersed in the wastewater changed significantly over the course of the wastewater treatment process. The taxonomic dissimilarity (Figure 4) observed among the immersed plant root biofilm samples showed three distinct communities: The anoxic (ANOX_B) and the closed aerobic (CLO_B); the first planted aerobic (HR1_B); the second and third planted aerobic tanks (HR2_B and HR3_B); and the sand filter (SAND_B). The changing characteristics of the first tanks of the treatment system (anoxic to aerobic conditions) is likely driving a transition from anaerobes such as Bacteriodes, and facultative anaerobes or microaerophiles such as Bacillus, Clostridium, Nocardia, Mycobacterium, to aerobic communities. The abundance of the facultative anaerobe group, Bacilli is likely promoted by the oxygen limiting environment in the first two tanks43. The immersed biofilm microbial community of the first planted aerobic tank (HR1) is taxonomically distinct from HR2 and HR3 despite identical physiochemical conditions. This pattern may be influenced by the composition of plant species in the individual tanks, which varies from tank to tank, or perhaps the diminishing influence of organisms contributed by feces.\n\nThere is limited ability to relate the results of our analysis of the microbial communities of this ecological WWTP to that of other systems serving different populations and geographic locations as this appears to be the first published findings on the topic. However, metagenomic analyses of the microbial ecology of conventional (activated sludge) wastewater systems have been described elsewhere59–62.\n\nFor example, Lee et al., 201459 employed 16S rRNA gene microarrays (PhyloChip) to establish a baseline microbial community structure of the municipal WWTP aeration basin. The microbial taxonomic composition of the aeration basin showed some similarities with that of the entire ecological WWTP sampled here. Specifically, Proteobacteria and Firmicutes were abundant in both the conventional system and ecological WWTP sampled here. To assess the microbial community seasonal variation of activated sludge over a four-year period Ju et al., 201463 employed WMS, as was used here. They showed variation in microbial taxonomic composition between the summer and winter samples. They also found variation in the microbial community composition over the four years sampled, irrespective of season. The metabolic structure of activated sludge according to SEED subgroups appeared to remain stable, in spite of variation in taxonomic composition, which suggests microbial community functional redundancy may be present in these systems. The ecological WWTP sampled for this work is housed in a climate-controlled glass house, which raises questions as to whether the microbial community varies from year to year and season to season. The microbial communities are not likely to exhibit temperature-dependent seasonal variation. Taxonomic variation may occur as a result of changes in the microbial communities contributed by the visitors.\n\nThe role plants have on influencing the structure of the root-colonizing microbial communities appeared to increase after the first three treatment tanks. The attenuation of taxa contributed by feces (Bacillus) after the first three treatment tanks is reflected in the increased microbial taxonomic heterogeneity in the latter two planted aerobic tanks (HR2 and HR3), which is reflected in the branch lengths for these samples in the neighbor joining network shown in Figure 4. These results indicate the need to design ecological WWTPs with sufficient retention time to allow for the attenuation of stool microbial communities and the development of diverse microbial biofilm communities.\n\nThe dominant PPCP removal process from the wastewater appears to be partitioning to sludge (biosolids) and biodegradation under nitrifying conditions, which are both reflected in the reductions in aqueous PPCP concentrations that occurred early in the treatment process. Primary sludge settles out of the wastewater in the holding tank and is periodically removed for off-site disposal. The concentration of some of the detected PPCPs continued to decline as the wastewater continued through the system (Table 1) indicating some continued removal beyond the first two or three tanks. For example, while the concentration of caffeine declined between the holding tank (9.5 × 104 ng L-1) and the anoxic tank (1.9 × 104 ng L-1), further reduction from the aqueous phase was observed in the subsequent three aerobic treatment tanks (CLO, HR1, HR2). Given that caffeine is a hydrophilic organic base (low KOW) only moderate partitioning to sludge is expected (86) and microbial biodegradation is likely to be responsible for the reduction in caffeine concentrations observed from the aerobic treatment tanks. Ibuprofen concentrations followed a pattern similar to caffeine’s in that significant reductions were seen in the first three aerobic treatment tanks (1.1 × 104 ng L-1 to 5.6 × 102 ng L-1). Carbamazepine, DEET, and trimethoprim concentrations remained stable throughout the treatment process. The combination of low partition to primary sludge expected and metabolic recalcitrance accounts for their stability in system47,50,55.\n\nMicrobial biodegradation pathways for most PPCPs have not been characterized, which makes it difficult to directly detect responsible genes. Nevertheless, ammonia oxidizing bacteria have been associated with the biodegradation of some PPCPs, while others have been shown to be degraded by nitrite oxidizing bacteria64,65. The relative abundances of sequences that MEGAN associated with ammonia monooxygenases were very low throughout the system (ranging from 0 to 53 reads), but were highest in the biofilm samples obtained from the HR3 and HR2 tanks. Due to lower dissolved oxygen levels, genes involved in denitrification (nitrite reductases) were found to be in highest relative abundance in the anoxic (ANOX) and closed (CLO) tanks (644 and 688, respectively).\n\nFunctional attributes of detected taxa reported in the literature can be used to identify metabolic potential pertinent to uncharacterized xenobiotic metabolic pathways. For example, in the first three tanks, the Firmicutes colonizing plant root surfaces have been reported to metabolize xenobiotics. Bacillus cereus, B. megaterium and B. amyloliquefaciens have been reported to metabolize phenol66, crude oil67, textile dyes68, and other xenobiotics through the induction of cytochrome P450s69,70. Dehalobacter sp. FTH1, identified in the plant root biofilm sample obtained from the second planted aerobic tank (HR2), has been reported to dechlorinate a number of organohalide xenobiotics71,72. Clostridium, identified in the root biofilm sample obtained from the first aerobic tank (HR1), has been reported to be involved in metabolism of bromophenols as a member of a consortium including Delhalobater73. Entercoccus spp., identified in the plant root biofilm of HR2, has been reported to degrade azo dyes74.\n\nOf the Actinomycetales identified, Mycobacterium spp., which have been reported to metabolize a variety of xenobiotics including polycyclic aromatic hydrocarbons75, biphenyls76, as well as various pharmaceuticals77 were identified in low relative abundance in samples obtained from the biofilm growing on plant roots in the anoxic tank and in high relative abundance in the biofilm sampled from sand filter. Mycobacterium spp. was also identified in the aqueous wastewater throughout the system (Figure 2 & Figure 9). While this metabolically plastic genus has been reported to be capable of metabolizing a wide variety of xenobiotics it should be noted that there are numerous pathogenic taxa including M. tuberculosis, M. bovis, and M. avium among others. The biofilm sample obtained from the sand filter contained 10,282 reads associated with human diseases and 18,766 reads associated with xenobiotic metabolism by the LCA algorithm in MEGAN.\n\nRhizobiales, which were identified in the aqueous phase throughout the system as well as in the biofilm sampled in the latter three treatment tanks (HR2, HR3, and SAND) have been reported as abundant in biofilm reactors treating sulfamethoxazole containing wastewater78. Also present throughout the system’s aqueous phase were members of the Rhodobacteraceae, Burkholderiaceae, Comamonadaceae, Pseudomonas, and Xanthomonadaceae, which have been reported to metabolize aromatic hydrocarbons79,80. Among these, the genus Pseudomonas has been identified as capable of biodegradation a variety of xenobiotic including some pharmaceuticals in a number of other settings including membrane bioreactors81, cultures originating from pharmaceutical wastewaters82, and environmental samples83.\n\nGreater xenobiotic metabolic heterogeneity was observed in the samples obtained from the plant root-associated biofilm as compared to the free-floating aqueous microbial community. The aqueous microbial metagenome, collectively, contained a greater total for xenobiotic metabolism gene copies (1.2 × 105 compared to 8.6 × 104 for the plant root biofilm) (see Supplementary material ST5). When comparing the proportion of sequences identified in the aqueous and biofilm phases of the system, which represent the non-root-associated and root-associated microbial populations, respectively, the xenobiotic metabolism gene categories nitrotoluene, benzoate, flurobenzoate and steroid degradation were found to be significantly higher in the aqueous phase samples (Welch’s t-test p-values < 0.05) (see Supplementary material ST49) (Figure 7). However, when comparing the type and abundance of reads associated with xenobiotic metabolism (KEGG level 3) the aqueous phase samples all resembled one another whereas, the biofilm samples were heterogeneous (see Figure 7 and Supplementary material ST31).\n\nFor aqueous phase microbial communities, growth of microbial consortia with the capacity to metabolize a given PPCP is driven by the presence of the compound in the wastewater, which will fluctuate with time. Stationary biofilm communities are likely to be more stable populations. These communities can accumulate with time and potentially acquire metabolic genes by horizontal gene transfer84. In contrast to the aqueous phase xenobiotic metabolism, which was dominated by benzoate degradation genes, the plant root microbial biofilm metagenomes contained higher relative abundances of other categories including aminobenzoate, chloroalkane/chloroalkene, and to a lesser degree ethylbenzene categories (Figure 7). However, of these categories, only chloroalakane/chloroalkene was significantly higher (Welches t-test p-value 0.031) in the biofilm samples collectively (see Supplementary material ST49).\n\nWhile others have identified ammonia oxygenases and nitrite reductases as being involved in microbial PPCP biodegradation64, the relative abundances of these gene categories were very low throughout the system. The relative abundance of xenobiotic metabolism gene copies was highest for the sand filter samples at 18,318 for SAND_W and 18,766 for SAND_B (see Supplementary material ST30). The sand filter’s ability to attenuate and accumulate sloughed off microbial cells as the wastewater passes through may be driving an accumulation of microbial biomass. If this is the case, sand filters are likely to have populations of the microbial communities found throughout the aqueous phase of the system, yet may not serve as a location of high metabolic activity, thereby contributing little to the metabolism of xenobiotics. The increase in concentrations observed for some of the PPCPs (caffeine, carbamazepine, DEET, gemfribozil, and ibuprofen) after the wastewater passed through the sand filter, if a real trend, could support this perspective.\n\nCulture bias was reflected in the taxonomic composition of the enrichment cultures growing on the sole pharmaceutical carbon sources examined here. While culture bias is well known85 and was expected, the enrichment of organisms capable of metabolizing individual pharmaceutical compounds from the WWTP effluent water reflects the ability for the ecological WWTP to support this metabolic capability. Only one taxon (Bacillus cereus) was identified in at least one enrichment culture (carbamazepine) and the biofilm sampled from the WWTP and one taxon (Mycobacterium spp.) was identified in the aqueous, biofilm and carbamazepine enrichment cultures (Figure 6). Given the selective pressure supplied by culturing, it is unlikely that these two taxa are solely responsible for the biodegradation of carbamazepine in the ecological WWTP. However, having isolated pharmaceutical metabolizing consortia from effluent water suggests the ecological WWTP supports microbial populations with the capacity remove recalcitrant micropollutants from wastewater.\n\nSecondary plant metabolites contributed to the wastewater by the tropical species cultivated in planted tanks may help support populations of microbial communities with xenobiotic degradation capabilities by providing a more consistent supply of structurally diverse carbon sources. The most abundant category of xenobiotic metabolism genes was associated with benzoate degradation in nearly all the samples. Benzoate degradation is known to play a role in the degradation of a variety of aromatic compounds23,86,87. The dominance of this functional category suggests that microbial communities in the aqueous phase were metabolizing a variety of benzoate-containing compounds, which were likely to include metabolites of plant and xenobiotic origin. However, further work is needed to determine whether plant-microbe feedback processes promote PPCP biodegradation in ecological WWTPs.\n\n\nData availability\n\nFASTQ files and associated metadata are available at NCBI BioProject ID PRJNA286671 (http://www.ncbi.nlm.nih.gov/bioproject/286671).",
"appendix": "Author contributions\n\n\n\nIB conceived of the study, designed the experiments and was awarded the funding that supported the work. HD contributed to the statistical and bioinformatics analysis. JV developed and refined the bioinformatics workflow as well as contributed bioinformatics analysis. ML created the enrichment cultures as a part of her senior thesis project. IB prepared the first full draft of the manuscript. IB and HD prepared the manuscript revisions. JV was involved in the revisions of the draft manuscript. All authors have contributed to and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors have no competing interests to declare.\n\n\nGrant information\n\nResearch reported in this publication was supported by an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences (NIGMS) of the National Institutes of Health (NIH) under grant number P20GM103449 awarded to INB through the Vermont Genetics Network (VGN). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NIGMS or NIH.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to acknowledge VGN for their efforts to promote research at undergraduate institutions. Timothy Hunter and Scott Tighe at the University of Vermont Advanced Genome Technologies Core (AGTC) developed and refined the protocol for, and completed, the genomic DNA extractions. Brewster Kingham at the University of Delaware, DNA Sequencing & Genotyping Center, Delaware Biotechnology Institute provided DNA sequencing. Finally, INB would like to acknowledge the guidance and feedback from Mary Tierney at the University of Vermont, Plant Biology Department and John Todd of Todd Ecological Design. INB would also like to acknowledge the numerous Lyndon State College undergraduate students that have contributed to this work through course work and senior thesis research. The authors are listed in the order of their contribution to this study.\n\n\nSupplementary material\n\nData tables ST1–49\n\nClick here to access the data.\n\nFigures S1–22\n\nClick here to access the data.\n\n\nReferences\n\nSingh KP, Mohan D, Sinha S, et al.: Impact assessment of treated/untreated wastewater toxicants discharged by sewage treatment plants on health, agricultural, and environmental quality in the wastewater disposal area. Chemosphere. 2004; 55(2): 227–255. PubMed Abstract | Publisher Full Text\n\nKolpin DW, Furlong ET, Meyer MT, et al.: Pharmaceuticals, hormones, and other organic wastewater contaminants in U.S. streams, 1999–2000: a national reconnaissance. Environ Sci Technol. 2002; 36(6): 1202–1211. PubMed Abstract | Publisher Full Text\n\nUnited Nations Educational, Scientific and Cultural Organization: World Water Assessment Programme, Water for People, Water for Life—the United Nations World Water Development Report. Berghahn Books: Barcelona, 2003. Reference Source\n\nTodd J, Brown EJ, Wells E: Ecological design applied. Ecol Eng. 2003; 20(5): 421–440. Publisher Full Text\n\nOdum HT: Environment, power, and society for the twenty-first century: the hierarchy of energy. New York: Columbia University Press. xiv, 2007; 418. Reference Source\n\nTodd J, Josephson B: The design of living technologies for waste treatment. Ecol Eng. 1996; 6(1–3): 109–136. Publisher Full Text\n\nMatamoros V, García J, Bayona JM: Organic micropollutant removal in a full-scale surface flow constructed wetland fed with secondary effluent. Water Res. 2008; 42(3): 653–660. PubMed Abstract | Publisher Full Text\n\nVymazal J: Horizontal sub-surface flow and hybrid constructed wetlands systems for wastewater treatment. Ecol Eng. 2005; 25(5): 478–490. Publisher Full Text\n\nLee CO, Howe KJ, Thomson BM: Ozone and biofiltration as an alternative to reverse osmosis for removing PPCPs and micropollutants from treated wastewater. Water Res. 2012; 46(4): 1005–1014. PubMed Abstract | Publisher Full Text\n\nReungoat J, Escher BI, Macova M, et al.: Biofiltration of wastewater treatment plant effluent: effective removal of pharmaceuticals and personal care products and reduction of toxicity. Water Res. 2011; 45(9): 2751–2762. PubMed Abstract | Publisher Full Text\n\nShrivastava P: Environmental technologies and competitive advantage. Strategic Manage J. 1995; 16(S1): 183–200. Publisher Full Text\n\nEtnier C, Guterstam Br: Ecological engineering for wastewater treatment. 2nd ed. Boca Raton: CRC Press. 1997; 451. Reference Source\n\nShao L, Wu Z, Zeng L, et al.: Embodied energy assessment for ecological wastewater treatment by a constructed wetland. Ecol Model. 2013; 252: 63–71. Publisher Full Text\n\nDaims H, Taylor MW, Wagner M: Wastewater treatment: a model system for microbial ecology. Trends Biotechnol. 2006; 24(11): 483–489. PubMed Abstract | Publisher Full Text\n\nTruu M, Juhanson J, Truu J: Microbial biomass, activity and community composition in constructed wetlands. Sci Total Environ. 2009; 407(13): 3958–3971. PubMed Abstract | Publisher Full Text\n\nPilon-Smits E: Phytoremediation. Annu Rev Plant Biol. 2005; 56: 15–39. PubMed Abstract | Publisher Full Text\n\nRovira AD: Plant root excretions in relation to the rhizosphere effect. Plant and Soil. 1956; 7(2): 178–194. Publisher Full Text\n\nChekol T, Vough LR, Chaney RL: Phytoremediation of polychlorinated biphenyl-contaminated soils: the rhizosphere effect. Environ Int. 2004; 30(6): 799–804. PubMed Abstract | Publisher Full Text\n\nDonnelly PK, Hegde RS, Fletcher JS: Growth of PCB-degrading bacteria on compounds from photosynthetic plants. Chemosphere. 1994; 28(5): 981–988. Publisher Full Text\n\nToussaint JP, Pham TT, Barriault D, et al.: Plant exudates promote PCB degradation by a rhodococcal rhizobacteria. Applied microbiology and biotechnology. 2012; 95(6): 1589–1603. PubMed Abstract | Publisher Full Text\n\nHaby PA, Crowley DE: Biodegradation of 3-chlorobenzoate as affected by rhizodeposition and selected carbon substrates. J Environ Qual. 1995; 25(2): 304–310. Publisher Full Text\n\nYi H, Crowley DE: Biostimulation of PAH degradation with plants containing high concentrations of linoleic acid. Environ Sci Technol. 2007; 41(12): 4382–4388. PubMed Abstract | Publisher Full Text\n\nSinger AC, Crowley DE, Thompson IP: Secondary plant metabolites in phytoremediation and biotransformation. Trends Biotechnol. 2003; 21(3): 123–130. PubMed Abstract | Publisher Full Text\n\nSinger AC, Wong CS, Crowley DE: Differential enantioselective transformation of atropisomeric polychlorinated biphenyls by multiple bacterial strains with different inducing compounds. Appl Environ Microbiol. 2002; 68(11): 5756–5759. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuttinun O, Müller R, Luepromchai E: Cometabolic degradation of trichloroethene by Rhodococcus sp. strain L4 immobilized on plant materials rich in essential oils. Appl Environ Microbiol. 2010; 76(14): 4684–4690. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerrer I, Zweigenbaum JA, Thurman EM: Analysis of 70 Environmental Protection Agency priority pharmaceuticals in water by EPA Method 1694. J Chromatogr A. 2010; 1217(36): 5674–5686. PubMed Abstract | Publisher Full Text\n\nEPA: Method 1697: Pharmaceuticals and Persona lCare Products in Water, Soil, Sediments, and Biosolids by HPLC/MS/MS. 2007; [cited 2016 6/13/2016]. Reference Source\n\nAtlas RM: Handbook of microbiological media. Fourth Edition, CRC press; 2010. Reference Source\n\nBolger AM, Lohse M, Usadel B: Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014; 30(15): 2114–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangmead B, Trapnell C, Pop M, et al.: Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009; 10(3): R25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao Y, Tang H, Ye Y: RAPSearch2: a fast and memory-efficient protein similarity search tool for next-generation sequencing data. Bioinformatics. 2012; 28(1): 125–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuchfink B, Xie C, Huson DH: Fast and sensitive protein alignment using DIAMOND. Nat Methods. 2015; 12(1): 59–60. PubMed Abstract | Publisher Full Text\n\nHuson DH, Mitra S, Ruscheweyh HJ, et al.: Integrative analysis of environmental sequences using MEGAN4. Genome Res. 2011; 21(9): 1552–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitra S, Gilbert JA, Field D, et al.: Comparison of multiple metagenomes using phylogenetic networks based on ecological indices. ISME J. 2010; 4(10): 1236–42. PubMed Abstract | Publisher Full Text\n\nOverbeek R, Begley T, Butler RM, et al.: The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res. 2005; 33(17): 5691–702. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanehisa M, Goto S, Sato Y, et al.: KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res. 2012; 40(Database issue): D109–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTatusov RL, Galperin MY, Natale DA, et al.: The COG database: a tool for genome-scale analysis of protein functions and evolution. Nucleic acids research. 2000; 28(1): 33–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoodall DW: A probabilistic similarity index. Nature. 1964; 203: 1098. Publisher Full Text\n\nBryant D, Moulton V: Neighbor-net: an agglomerative method for the construction of phylogenetic networks. Mol Biol Evol. 2004; 21(2): 255–265. PubMed Abstract | Publisher Full Text\n\nParks DH, Tyson GW, Hugenholtz P, et al.: STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics. 2014; 30(21): 3123–3124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrzywinski M, Schein J, Birol I, et al.: Circos: an information aesthetic for comparative genomics. Genome Res. 2009; 19(9): 1639–1645. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCenter VAoAI: Visitor Traffic Statistics. 2013; [cited 2016 6/13/2016]. Reference Source\n\nDrobniewski FA: Bacillus cereus and related species. Clin Microbiol Rev. 1993; 6(4): 324–338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXia K, Bhandari A, Das K, et al.: Occurrence and fate of pharmaceuticals and personal care products (PPCPs) in biosolids. J Environ Qual. 2005; 34(1): 91–104. PubMed Abstract | Publisher Full Text\n\nSuárez S, Carballa M, Omil F, et al.: How are pharmaceutical and personal care products (PPCPs) removed from urban wastewaters? Reviews in Environmental Science and Bio/Technology. 2008; 7(2): 125–138. Publisher Full Text\n\nDobbs RA, Wang L, Govind R: Sorption of toxic organic compounds on wastewater solids: correlation with fundamental properties. Environ Sci Technol. 1989; 23(9): 1092–1097. Publisher Full Text\n\nCarballa M, Fink G, Omil F, et al.: Determination of the solid-water distribution coefficient (Kd) for pharmaceuticals, estrogens and musk fragrances in digested sludge. Water Res. 2008; 42(1–2): 287–295. PubMed Abstract | Publisher Full Text\n\nOulton RL, Kohn T, Cwiertny DM: Pharmaceuticals and personal care products in effluent matrices: A survey of transformation and removal during wastewater treatment and implications for wastewater management. J Environ Monit. 2010; 12(11): 1956–1978. PubMed Abstract | Publisher Full Text\n\nLishman L, Smyth SA, Sarafin K, et al.: Occurrence and reductions of pharmaceuticals and personal care products and estrogens by municipal wastewater treatment plants in Ontario, Canada. Sci Total Environ. 2006; 367(2–3): 544–558. PubMed Abstract | Publisher Full Text\n\nCarballa M, Omil F, Lema JM, et al.: Behavior of pharmaceuticals, cosmetics and hormones in a sewage treatment plant. Water Res. 2004; 38(12): 2918–2926. PubMed Abstract | Publisher Full Text\n\nShon H, Vigneswaran S, Snyder S: Effluent organic matter (EfOM) in wastewater: constituents, effects, and treatment. Crit Rev Environ Sci Technol. 2006; 36(4): 327–374. Publisher Full Text\n\nBoreen AL, Arnold WA, McNeill K: Photodegradation of pharmaceuticals in the aquatic environment: a review. Aquatic Sciences. 2003; 65(4): 320–341. Publisher Full Text\n\nBoreen AL, Arnold WA, McNeill K: Photochemical fate of sulfa drugs in the aquatic environment: sulfa drugs containing five-membered heterocyclic groups. Environ Sci Technol. 2004; 38(14): 3933–3940. PubMed Abstract | Publisher Full Text\n\nOnesios KM, Yu JT, Bouwer EJ: Biodegradation and removal of pharmaceuticals and personal care products in treatment systems: a review. Biodegradation. 2009; 20(4): 441–466. PubMed Abstract | Publisher Full Text\n\nPérez S, Eichhorn P, Aga DS: Evaluating the biodegradability of sulfamethazine, sulfamethoxazole, sulfathiazole, and trimethoprim at different stages of sewage treatment. Environ Toxicol Chem. 2005; 24(6): 1361–1367. PubMed Abstract | Publisher Full Text\n\nKarnjanapiboonwong A, Morse AN, Maul JD, et al.: Sorption of estrogens, triclosan, and caffeine in a sandy loam and a silt loam soil. J Soils Sediments. 2010; 10(7): 1300–1307. Publisher Full Text\n\nBatt AL, Kim S, Aga DS: Comparison of the occurrence of antibiotics in four full-scale wastewater treatment plants with varying designs and operations. Chemosphere. 2007; 68(3): 428–435. PubMed Abstract | Publisher Full Text\n\nSegata N, Haake SK, Mannon P, et al.: Composition of the adult digestive tract bacterial microbiome based on seven mouth surfaces, tonsils, throat and stool samples. Genome Biol. 2012; 13(6): R42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee S, Geller JT, Torok T, et al.: Characterization of wastewater treatment plant microbial communities and the effects of carbon sources on diversity in laboratory models. PLoS One. 2014; 9(8): e105689. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcLellan SL, Huse SM, Mueller-Spitz SR, et al.: Diversity and population structure of sewage‐derived microorganisms in wastewater treatment plant influent. Environ Microbiol. 2010; 12(2): 378–392. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShanks OC, Newton RJ, Kelty CA, et al.: Comparison of the microbial community structures of untreated wastewaters from different geographic locales. Appl Environ Microbiol. 2013; 79(9): 2906–2913. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang Y, Yu K, Xia Y, et al.: Metagenomic analysis of sludge from full-scale anaerobic digesters operated in municipal wastewater treatment plants. Appl Microbiol Biotechnol. 2014; 98(12): 5709–5718. PubMed Abstract | Publisher Full Text\n\nJu F, Guo F, Ye L, et al.: Metagenomic analysis on seasonal microbial variations of activated sludge from a full‐scale wastewater treatment plant over 4 years. Environ Microbiol Rep. 2014; 6(1): 80–89. PubMed Abstract | Publisher Full Text\n\nFernandez-Fontaina E, Gomes IB, Aga DS, et al.: Biotransformation of pharmaceuticals under nitrification, nitratation and heterotrophic conditions. Sci Total Environ. 2016; 541: 1439–1447. PubMed Abstract | Publisher Full Text\n\nFernandez-Fontaina E, Omil F, Lema JM, et al.: Influence of nitrifying conditions on the biodegradation and sorption of emerging micropollutants. Water Res. 2012; 46(16): 5434–5444. PubMed Abstract | Publisher Full Text\n\nBanerjee A, Ghoshal AK: Phenol degradation by Bacillus cereus: pathway and kinetic modeling. Bioresour Technol. 2010; 101(14): 5501–5507. PubMed Abstract | Publisher Full Text\n\nDas K, Mukherjee AK: Crude petroleum-oil biodegradation efficiency of Bacillus subtilis and Pseudomonas aeruginosa strains isolated from a petroleum-oil contaminated soil from North-East India. Bioresour Technol. 2007; 98(7): 1339–1345. PubMed Abstract | Publisher Full Text\n\nYe J, Singh A, Ward OP: Biodegradation of nitroaromatics and other nitrogen-containing xenobiotics. World J Microbiol Biotechnol. 2004; 20(2): 117–135. Publisher Full Text\n\nLončar N, Božić N, Lopez-Santin J, et al.: Bacillus amyloliquefaciens laccase--from soil bacteria to recombinant enzyme for wastewater decolorization. Bioresour Technol. 2013; 147: 177–183. PubMed Abstract | Publisher Full Text\n\nNarhi LO, Fulco AJ: Identification and characterization of two functional domains in cytochrome P-450BM-3, a catalytically self-sufficient monooxygenase induced by barbiturates in Bacillus megaterium. J Biol Chem. 1987; 262(14): 6683–6690. PubMed Abstract\n\nYoshida N, Ye L, Baba D, et al.: A novel Dehalobacter species is involved in extensive 4,5,6,7-tetrachlorophthalide dechlorination. Appl Environ Microbiol. 2009; 75(8): 2400–2405. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYoshida N, Ye L, Baba D, et al.: Reductive Dechlorination of Polychlorinated Biphenyls and Dibenzo-p-Dioxins in an Enrichment Culture Containing Dehalobacter Species . Microbes Environ. 2009; 24(4): 343–346. PubMed Abstract | Publisher Full Text\n\nLi Z, Yoshida N, Wang A, et al.: Anaerobic mineralization of 2,4,6-tribromophenol to CO2 by a synthetic microbial community comprising Clostridium, Dehalobacter, and Desulfatiglans. Bioresour Technol. 2015.176: 225–232. PubMed Abstract | Publisher Full Text\n\nPunj S, John GH: Purification and identification of an FMN-dependent NAD(P)H azoreductase from Enterococcus faecalis. Curr Issues Mol Biol. 2009; 11(2): 59–65. PubMed Abstract\n\nCerniglia CE: Biodegradation of polycyclic aromatic hydrocarbons. Biodegradation. 1992; 3(2–3): 351–368. Publisher Full Text\n\nMoody J, Doerge DR, Freeman JP, et al.: Degradation of biphenyl by Mycobacterium sp. strain PYR-1. Appl Microbiol Biotechnol. 2002; 58(3): 364–369. PubMed Abstract | Publisher Full Text\n\nBesse P, Combourieua B, Poupin P, et al.: Degradation of morpholine and thiomorpholine by an environmental Mycobacterium involves a cytochrome P450. Direct evidence of intermediates by in situ 1H NMR. J Mol Catal B Enzym. 1998; 5(1–4): 403–409. Publisher Full Text\n\nEsplugas M, González O, Sans C: Bacterial community characterization of a sequencing batch reactor treating pre-ozonized sulfamethoxazole in water. Environ Technol. 2013; 34(9–12): 1583–1591. PubMed Abstract | Publisher Full Text\n\nHerbst FA, et al.: Elucidation of in situ polycyclic aromatic hydrocarbon degradation by functional metaproteomics (protein-SIP). Proteomics. 2013; 13(18–19): 2910–2920. PubMed Abstract | Publisher Full Text\n\nJeon CO, Madsen EL: In situ microbial metabolism of aromatic-hydrocarbon environmental pollutants. Curr Opin Biotechnol. 2013; 24(3): 474–481. PubMed Abstract | Publisher Full Text\n\nDe Gusseme B, Vanhaecke L, Verstraete W, et al.: Degradation of acetaminophen by Delftia tsuruhatensis and Pseudomonas aeruginosa in a membrane bioreactor. Water Res. 2011; 45(4): 1829–1837. PubMed Abstract | Publisher Full Text\n\nShourian M, Noghabi KA, Zahiri HS, et al.: Efficient phenol degradation by a newly characterized Pseudomonas sp. SA01 isolated from pharmaceutical wastewaters. Desalination. 2009; 246(1–3): 577–594. Publisher Full Text\n\nBenotti MJ, Brownawell BJ: Microbial degradation of pharmaceuticals in estuarine and coastal seawater. Environ Pollut. 2009; 157(3): 994–1002. PubMed Abstract | Publisher Full Text\n\nTop EM, Springael D: The role of mobile genetic elements in bacterial adaptation to xenobiotic organic compounds. Curr Opin Biotechnol. 2003; 14(3): 262–269. PubMed Abstract | Publisher Full Text\n\nStaley JT, Konopka A: Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Annu Rev Microbiol. 1985; 39(1): 321–346. PubMed Abstract | Publisher Full Text\n\nvan der Meer JR, de Vos WM, Harayama S, et al.: Molecular mechanisms of genetic adaptation to xenobiotic compounds. Microbiol Rev. 1992; 56(4): 677–694. PubMed Abstract | Free Full Text\n\nDíaz E: Bacterial degradation of aromatic pollutants: a paradigm of metabolic versatility. Int Microbiol. 2004; 7(3): 173–180. PubMed Abstract"
}
|
[
{
"id": "16355",
"date": "27 Sep 2016",
"name": "Benjamin C. Kirkup",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn reviewing “Metagenomic analysis of an ecological wastewater treatment plant’s microbial communities and their potential to metabolize pharmaceuticals,” I consider the intended breadth of the paper. The paper would accurately be described as a first look at a single ‘ecological wastewater treatment plant’ through the somewhat particular lenses of metagenomics and micropollutants.\nMicropollutants are not the primary target of wastewater treatment; prevention of infectious disease motivated the first sand filters and avoiding eutrophication or other bulk pollution is a common goal, which explains the focus on nitrogen, dissolved organic carbon, phosphorus, etc. Micropollutants are not successfully managed by most wastewater treatment paradigms, as the authors discuss. The authors integrated a number of methods to characterize micropollutants and communities in small-scale wastewater treatment facility that integrates plants and microbes into the treatment process. In general, the authors seem to be advocates for the special treatment process; the intention is to evaluate the processing of these recalcitrant pollutants; but no direct comparison is made to other processes.\nA single facility was sampled across the process at a single time. Input wastewater and waste in process was sampled and the influent and resident micropollutants were quantified.\nThe enrichment cultures were performed with each of three pharmaceuticals as the sole carbon source, in replicate. Of 15 replicates, eight were selected based on visual inspection for microbial growth.\n\nMethods and Results:\nI do not have the background to critique the micropollutant quantification, which was done at a commercial lab using standard EPA methods.\n\nThere are substantial problems with the sequencing methods. The shotgun metagenomics was done without a kit (true negative) control or a positive (synthetic) community standard. In addition, no standards were used to assess DNA extraction biases (such as reported by13. The EZNA Mollusc DNA kit was used; I am not sufficiently familiar with this specific kit, compared to the bacterial kits, for example. This is the only case in which that kit is used in bacterial shotgun metagenomics, to my knowledge. I do not know what kinds of bias this introduces to the analysis. Overall, the use of non-standard methods requires new controls; some of which should be performed even when standard methods are employed. Comparisons to non-sequencing methods (ie. microscopy, flow cytometry, qPCR from several DNA extractions) would also inform evaluation of bias and uncertainty.\nThe bioinformatic methods reveal problems interpreting the data. MEGAN is the core of the bioinformatics pipeline, which was tested on two existing synthetic datasets. ~11-12% of the reads were excluded for quality. 50-60% of the remaining reads were protein reads from NCBI. However, environmental and enrichment culture reads differed greatly in their assignment to known taxonomy. Almost all enrichment culture reads were assigned; only 80% of the quality and protein reads from the environment could be assigned.\n\nEukaryotes account for a substantial fraction of the reads in HR2_B, but not in the others samples. Fungi are specifically reported in the enrichment cultures, suggesting that eukaryotes are indeed sequenced and analyzed. Given all this information, the lack of plant DNA in the other HR samples is counter-intuitive. Typically, host-associated shotgun metagenomic sequencing effort would benefit from draft host genome sequencing, to allow the host DNA to be removed bioinformatically (the way it was with the PhiX spike-in). Alternative (wet) methods to remove host DNA are also not evident.\n\nThe reported microbial community compositions are unexpectedly simple. Perhaps the depth of sampling was inadequate; this could be assessed with a method such as 16S rRNA sequencing at a greater depth. The comparison of ‘taxonomic,’ ‘community,’ or ‘metabolic’ structures is undermined by the need to map reads to the KEGG database without assembly and with a very low rate of read mapping to NCBI proteins. Further, rarefaction (normalizing the total number of reads) is problematic2. Given the difficulties with data analysis, however, it seems unnecessary to give a detailed critique of the downstream analyses.\nSome organisms were present in the original samples which apparently survived in the enrichment culture. Some of these have been identified by sequencing, but it is unclear how many more may be present. No attempt was made to return to the original samples with probes for the organisms found during enrichment, nor were genomes assembled. With a significantly greater number of enrichments, multi-sample strategies might have been used to assemble genomes from distinct organisms4. These could be used to map reads from the original samples.\nConclusions:\nGiven that so few samples were examined, the study is best scaled to pilot a larger examination of similar treatment facilities. Samples were not studied in technical replicate, nor repeated in space across the tanks, nor across time at short and long time scales, or across related facilities, or facilities with other processing strategies. No experiments were performed which perturbed or altered the plant or microbial processes in the facility.\nAs a result, most of the conclusions must be drawn with very limited statistical evaluation. Most of the conclusions are not effectively supported because the sample variance could not be estimated. Conclusions like “The aqueous microbial metagenome, collectively, contained a greater total for xenobiotic metabolism gene copies…” are unsupported.\n\nA starting hypothesis is that the fraction of reads from a sample assigned to xenobiotic degradation is indistinguishable from sample to sample, rather than being ordered (“produced the greatest number of sequences…”). Similarly, statements about the taxonomy (\"relative abundance of Bacilli in the immersed biofilm communities was highly variable\") are unsupported by the data.\nThe \"role of plants\" (\"The role plants have on influencing the structure of the root-colonizing microbial communities appeared to increase after the first three treatment tanks\") could not be examined by single observations of the root and water within individual tanks. Nor could conclusions be reached about microbial community variation over time (see: “are likely to change with time, fluctuating…”), given that only a single time point was sampled. The results do not indicate that “the ecological WWTP is performing well with regard to attenuating fecal coliforms” given that infection can be caused by a low number of cells and the sensitivity of the sequencing to minor populations was uncharacterized.\nThe discussion of phase partitioning of micropollutants (\"PPCPs in the wastewater\" and \"Removal of PPCPs by the ecological WWTP\") is outside my area of expertise.\nGeneralization of microbial behavior at the level of species is problematic, particularly when inferring activity on complex organic substrates. Generalizing to families and phyla (Rhizobiales, Firmicutes, Mycobacteria... ) is inappropriate. Experiments with specific strains are required for effective experimental annotation of novel degradation pathways.\nOverall Review: This is an interesting pilot experiment for studies which collectively may result in conclusions of the sort reported in the paper. Method development and the sequencing of specific microbial and plant reference genomes will prepare a team for the extensive sampling and experimentation needed to assess the effectiveness and impact of this wastewater treatment strategy. The concentrations of micropollutants observed in these few samples suggest that they can be altered by such treatment, but that they also may circulate at increased concentrations as water is reused.",
"responses": [
{
"c_id": "2205",
"date": "28 Sep 2016",
"name": "Ian Balcom",
"role": "Author Response",
"response": "Thank you for the thoughtful review of our work and helpful feedback. We will be incorporating many of your suggestions into future studies, which will be focusing on increasing replication. Here are responses to some of your comments on the extraction methods used here: The Modified Omega Mollusc Kit protocol used in this study was modified to act as a pre -packaged CTAB kit described J Zhou and Ed Moore; DNA recovery from soils of diverse composition. Appl Environ Microbiol. 1996 Feb;62(2):316-22. Zhou J, Bruns MA, Tiedje JM. Simplified protocols for the preparation of genomic DNA from bacterial cultures. Molecular Microbial Ecology Manual, 2nd Edition 1.01: 3–18, 2004. E. Moore, A Arnscheidt, A. Kruger, C. Strompl, M. Mau This CTAB protocol has also been described in the following manuscript with the following modifications: A pre-extraction metagenomics bacterial cell wall digestion was performed using a multi-component enzyme mix described below in the MetaSub paper. This enzyme is manufactured by Sigma and will be commercially available in Q1 2017. The Metagenomics and Metadesign of the Subways and Urban Biomes (MetaSUB) International Consortium inaugural meeting report. The MetaSUB International Consortium. Microbiome. 2016 4:24 This Reference data and method is also available on the following websites: https://abrf.org/sites/default/files/temp/RGs/MGRG/narg-meta-2013_final_poster_forprinter.pptx http://extrememicrobiome.org/links/ Comments on DNA kit Negative and Library Negatives: Our core facility runs DNA extraction kit negatives on the lot of reagents from both Sigma and Omega BioTek as a \"process control\" and quantifies the negative using the Qubit High Sensitivity kit. We do not sequence negative controls that have values that are below detection limit [BDL]. A Negative control is processed for each group of DNA library samples. Comments on DNA extraction bias and microbial standards: Currently there is no approved microbial whole cell reference standards. However, the ABRF Metagenomics research group is currently in the process of distributing 6 through a collaboration with ATCC. The ABRF is the first group to make a whole cell microbial standard. Further the ABRF has assembled the Class 1 and Class1+ genomic microbial standard that will also be distributed through the ATCC. This topic was discussed at great length at the NIST Microbiome standards meeting in Gaitherburg MD on Aug 8th as communicated by Scott Tighe. Scott Tighe is the current lab manager that performed this work, is one of four co-leaders of the NIST International Microbiome Metagenomics Standards Alliance [IMMSA] and presented these bottle necks on day 2 session 3. It is well recognized that whole cell quantitative standards do not exist; therefore, the reviewer must understand that doing DNA extraction bias studies are a large undertaking. The ABRF-ATCC whole cell microbial Class 1 reference standard will help us understand the DNA biases in the the entire field of microbiome and not just rely on a simple MoBio protocol. It should be pointed out to the reviewer that Zymo's Zymobiotics bacterial standard is not whole cell, but rather suspended in SDS as a preservative and can not be used for DNA extraction bias studies. https://www.nist.gov/news-events/events/2016/08/standards-microbiome-measurements-workshop"
}
]
},
{
"id": "17273",
"date": "14 Nov 2016",
"name": "Andrew C. Singer",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI found this to be an interesting field-based study. It is limited by all the typical factors that one finds in a field-based study (i.e., replication, no controls, etc), but that is to be expected and okay so long as the analysis and interpretation are consistent with theses limitations. I feel the authors largely stayed true to their experimental design and offered some useful insights into the processes of an ecological WWTP. I only have a few trivial suggestions to for (potentially) improving some elements of the paper, although it reads well and could largely remain as it stands.\nSpecific suggestions: Would it be appropriate to denote \"ecological wastewater treatment plants\" as eWWTP, as they are clearly different from WWTPs in this context.\nThe aim of the study does not immediately follow from the sentence in the abstract: \"To determine whether the removal of micropollutants in ecological wastewater treatment plants (WWTPs) is promoted by the plant-microbe interactions...\"\nYou are assessing the correlation between wastewater input, treatment stage, PPCP removal, bacterial family and gene prevalence in plant and sand biofilms, planktonic cells and enrichment cultures.\nGranted there's no easy way to summarize this, but it is a more accurate description of the paper . You may wish to adjust the abstract to better reflect what was done.\nSampling: It's not clear how samples were collected from the eWWTP, the replication or pooling that was done, the transportation conditions and storage. Clarify when sampling was of water and/or biofilm and depth of sample acquisition and whether roots were removed or scraped for the biofilm (and the sand filter?). The environmental conditions in the eWWTP need to be provided (temp, pH, atmosphere, etc).\nThe observation that biodegradation/metabolism is dominated by benzoate degradation is quite interesting and leads to some useful insight into what might be going on in the eWWTP. It might be an option to make this a focus of the paper, i.e., in the title. If you choose to do this you might then rearrange the results to put this first as the sequence information really just supports this one result.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1881
|
https://f1000research.com/articles/5-1479/v1
|
23 Jun 16
|
{
"type": "Software Tool Article",
"title": "SNPsplit: Allele-specific splitting of alignments between genomes with known SNP genotypes",
"authors": [
"Felix Krueger",
"Simon R. Andrews",
"Simon R. Andrews"
],
"abstract": "Sequencing reads overlapping polymorphic sites in diploid mammalian genomes may be assigned to one allele or the other. This holds the potential to detect gene expression, chromatin modifications, DNA methylation or nuclear interactions in an allele-specific fashion. SNPsplit is an allele-specific alignment sorter designed to read files in SAM/BAM format and determine the allelic origin of reads or read-pairs that cover known single nucleotide polymorphic (SNP) positions. For this to work libraries must have been aligned to a genome in which all known SNP positions were masked with the ambiguity base ’N’ and aligned using a suitable mapping program such as Bowtie2, TopHat, STAR, HISAT2, HiCUP or Bismark. SNPsplit also provides an automated solution to generate N-masked reference genomes for hybrid mouse strains based on the variant call information provided by the Mouse Genomes Project. The unique ability of SNPsplit to work with various different kinds of sequencing data including RNA-Seq, ChIP-Seq, Bisulfite-Seq or Hi-C opens new avenues for the integrative exploration of allele-specific data.",
"keywords": [
"Allele-specific",
"SNP",
"N-masking",
"ASM",
"ASE",
"ASB",
"Allele"
],
"content": "Introduction\n\nMost functional NGS studies performed today still ignore the fact that many model organisms are diploid, and work on the averaged signal from the two alleles. However, a complete understanding of the biology of diploid organisms requires that the two alleles be measured separately. Allele-specific analysis of next-generation sequencing reads is becoming an important tool to identify events such as allele-specific expression of genes (ASE), allele-specific binding of transcription factors or histones (ASB) or allele-specific methylation (ASM). These techniques allow a more detailed investigation of the effects of genetic or epigenetic variation on genome regulation or studying parent of origin effects such as genomic imprinting or allelic imbalance.\n\nThere are two main use cases for the investigation of allele-specific events: If both parental genotypes are clean and known in advance, e.g. for defined crosses of inbred mouse strains, parent of origin specific effects can be studied by comparing the two parental genotypes. Alternatively, allele-specific analyses require the more complex procedure of whole genome haplotype reconstruction (e.g. as described in 1). For the purposes of this manuscript we will use the terms ‘Allele 1’ or ‘Allele 2’ to refer to the maternal or paternal genotype, respectively, or to a reference and alternative strain or genome if the distinction between maternal/paternal is not meaningful.\n\nThe detection of allele-specific events relies on the ability to distinguish the two alleles of a diploid organism, which can be accomplished by looking at reads covering heterozygous single nucleotide polymorphisms (SNPs), small insertions or deletions (InDels) or greater structural variations. While the allele-specific analysis of InDels has been found to be challenging2, the use of SNPs to discriminate alleles is the most widely used approach because it allows for the maintenance of a common set of reference genome coordinates.\n\nSeveral approaches have been taken to perform allele-specific alignments. The simplest is to align all reads to a single reference genome, but this introduces a bias as reads from the allele which is more similar to the reference are able to map more efficiently3. Another approach involves the generation of two personalised genomes by incorporating known SNP positions (and possibly InDels) followed by an alignment to both genomes and finally a post-processing step to compute the union of the separate alignments (used in different flavours in 4–6). This approach is slower as it requires two separate mapping steps, and can still result in allelic bias because reads from one allele might not map uniquely or to an incorrect location in one of the genomes3. A more recent improvement7 aims to reduce mapping biases by realigning reads that overlap SNP positions in all possible allele combinations and keeping only reads that align to the same position in both genomes - this reduces bias, but is computationally complex. Finally, the issue of bias can be tackled by masking polymorphic sites with the ambiguity nucleobase ‘N’ (henceforth called ‘N-masking’), performing a single alignment to the N-masked genome and then assigning reads based on the sequence found underneath the masked positions. The rationale for N-masking in allele-specific alignments is that the mapping bias towards the reference allele is eliminated and both alleles of the same read get placed in the same position in the genome equally well. N-masking the genome is a one-off exercise and this approach has the advantage of requiring only a single alignment to a reference which noticeably reduces the computational load. Despite the fact that N-masking effectively avoids allelic biases it may occasionally result in a minor loss of sensitivity when the density of N covered by a read is getting too high.\n\nA requirement for N-masking is that SNP positions are known, e.g. via a public resource like the Mouse Genomes Project which provides high quality variant calls for a large number of mouse strains8 (hosted at http://www.sanger.ac.uk/science/data/mouse-genomes-project). If the genotype is not known, SNP positions may be called from the data itself, or from genome re-sequencing performed in parallel. The quality of the genotype calls is crucial for allele assignment, so the genotype data needs to be collected carefully and quality control and filtering is required to avoid biases and false positive hits9. Further downstream analysis of allele-specific data is highly dependent on the experiment type and is beyond the scope of this manuscript.\n\nTo our knowledge there are currently no user-friendly solutions available for the allele-specific splitting of sequencing reads aligned to N-masked genomes. We sought to address this by creating SNPsplit, an easy-to-use tool for assigning allele-specific reads. In its generic mode SNPsplit is not tied to any particular aligner and operates across several different experiment types including RNA-Seq, genomic DNA-alignments, DNA methylation (Bisulfite-Seq) and 3-D genome organisation (Hi-C). While a similar allele-specific functionality has been integrated into specialised applications, e.g. HiC-Pro10, the unique capability to work with several different data types renders SNPsplit an ideal choice for correlation studies using allele-specific sequencing reads.\n\n\nMethods\n\nSNPsplit is written in Perl and consists of three separate scripts that can be run individually on the command line. It takes alignment files in BAM/SAM format as input and further requires an annotation file containing the positions of all SNPs in the genome. SNPsplit determines for each aligned read whether it overlaps with a known SNP position and adds a tag to the alignment that indicates whether the read can be assigned to a specific allele or is unassignable. The reads are then sorted into different sub-files depending on the library type, i.e. single-end or paired-end, and the nature of the sample, e.g. RNA-Seq, BS-Seq or Hi-C.\n\nAs long as a SAM/BAM file that was aligned to an N-masked genome is provided as input SNPsplit should perform well regardless of how the N-masking itself was accomplished. Since there is an ever growing number of genomes and different SNP annotation files and file formats it would be too much to ask to provide a generally applicable way of constructing N-masked genomes that fits all cases.\n\nWe do however provide an automated solution to generate N-masked versions of the genome for all strains in the Mouse Genomes Project (http://www.sanger.ac.uk/science/data/mouse-genomes-project). The genome preparation step supports the generation of single hybrid strains where one allele is the same as the mouse reference sequence (which is based on strain C57BL/6J, hereafter called Black 6) and one alternative allele, e.g. SPRET/EiJ. It also supports the generation of dual hybrid strains where both alleles are different from the Black 6 reference, e.g. CAST/EiJ and 129S1/SvImJ. At the time of writing the Mouse Genomes Project encompassed variation information for 36 different mouse strains; the SNP annotation data for all strains relative to Black 6 reference sequence may be found in the variant call format (VCF) file ‘mgp.v5.merged.snps_all.dbSNP142.vcf.gz’ (VCF v4.2; last modified 13 May 2015; download available at: ftp://ftp-mouse.sanger.ac.uk/current_snps/). The SNPsplit genome preparation first reads the SNP annotations for the strain in question from the VCF file and then constructs the N-masked genomes based on the Black 6 reference sequence using only high confidence homozygous positions. The process is slightly different for single or dual hybrid strains.\n\nSingle-Hybrid Strains. This generates a new genome sequence, with SNPs either N-masked or included as full sequence, where Allele 1 (or Genome 1) is the Black 6 reference and Allele 2 (or Genome 2) is the alternative strain.\n\n1) The VCF file is read and filtered for high-confidence SNPs for the strain specified\n\n2) The Black 6 reference genome is read into memory, and the filtered high-confidence SNP positions are incorporated either as N-masking (default) or full sequence (optional)\n\nDual-Hybrid Strains. This generates a new genome sequence where neither allele is the Black 6 reference. SNPs can be either N-masked or included as full sequence, where Allele 1 (or Genome 1) is the strain specified as strain 1 and Allele 2 (or Genome 2) is the strain specified as strain 2.\n\n1) The VCF file is read and filtered for high-confidence SNPs in strain 1\n\n2) The Black 6 reference genome is read into memory, and the filtered high-confidence SNP positions are incorporated as full sequence and N-masking (optional)\n\n3) The VCF file is read and filtered for high-confidence SNPs in strain 2\n\n4) The filtered high-confidence SNP positions of strain 2 are incorporated as full sequence and N-masking (optional)\n\n5) The SNP information of strain 1 and strain 2 relative to the Black 6 reference genome build are compared and a new Ref/SNP annotation is constructed whereby the new Ref/SNP information will be strain 1/strain 2\n\n6) The full genome sequence of strain 1 is read into memory, and the high-confidence SNP positions between strain 1 and strain 2 are incorporated as full sequence and N-masking (optional)\n\nThe N-masked sequences (or sequences containing the full sequence SNPs) are written out in FASTA format and ready to be indexed with the alignment software of your choice. Alignments to N-masked genomes are not very different to regular mapping except that they require the aligner to support ambiguity DNA bases such as N. Software confirmed to be working for this approach include (but are not limited to) Bowtie211, BWA12, HISAT213, STAR14 or any tool wrapping one of these aligners.\n\nSNPsplit operates in two stages which are run sequentially: I) read tagging and II) read sorting. Both steps generate detailed reports for record keeping.\n\nStage I: Tagging SNPsplit analyses reads for overlaps with known SNP positions for which it requires the mismatch position field (MD:Z:) in the SAM entry, and writes out a tagged BAM file in the same order as the original file. This process requires a list of all known SNP positions between the two different genomes (supplied as a SNP file) and works on a read-by-read basis.\n\nRead tagging generally works as a multi-step process:\n\n1. Determine the position(s) in the read that overlap genomic N(s)\n\n2. Adjusting position for insertions/deletions\n\n3. Determine equivalent genomic position\n\n4. Determine if the SNP is present in the list of SNP positions, and if yes whether the position in the read was the Allele 1 or Allele 2 base\n\nDepending on the collected SNP information the tagging module then determines whether a read can be assigned to a certain allele and appends an additional optional field ‘XX:Z:tag’ to the SAM entry of each read. The tag can be one of the following:\n\n• UA - Unassigned\n\n• G1 - Genome 1-specific (Allele 1, the reference)\n\n• G2 - Genome 2-specific (Allele 2, the alternative strain)\n\n• CF - Conflicting\n\nReads are considered unassignable (UA) if they do not overlap any known SNP position. Reads harbouring at least one SNP specific for both genomes at the same time are classified as conflicting (CF).\n\nThe determination of overlaps is geared to handle the CIGAR operations M (match to the reference), D (deletion in the read), I (insertion in the read) and N (skipped regions, used for splice mapping). Other CIGAR operations (see the SAM format specification for further details15) are currently not supported. This means that SNPsplit requires reads to be a full match from end-to-end and thus soft-clipping (CIGAR operation: S), which may introduce artefactual alignments to poorly annotated regions in the genome16 is not supported (see also section Use Cases for RNA-Seq below on how to avoid soft-clipping issues).\n\nStage II: Sorting The tagged BAM file is read in again and sorted into allele-specific files according to their XX:Z: tag. For paired-end or Hi-C experiments the combination of tags for both Read 1 and Read 2 are considered (see below for examples). Conflicting reads, or also disagreeing read-pairs for paired-end samples, are not printed out by default. The sorting process may also be run stand-alone on tagged BAM files to try out different sorting options (e.g. separating out paired-end and singleton alignments or enabling reporting of conflicting alignments).\n\nSNPsplit runs on any Linux-based operating system with Perl installed (tested using CentOS v6.2 and Perl v5.10.1). In addition, a functional version of SAMtools15 (v0.1.18 or later) is required for handling of SAM/BAM files. Memory requirements depend directly on the genome size and the total number of heterozygous SNPs to be stored, but as a guideline 5–10 GB RAM should be sufficient to process data for most mouse strains.\n\n\nUse cases\n\nSNPsplit is able to handle any kind of standard genomic alignment file irrespective of the method employed to generate the library as long as the CIGAR operation requirements are met (see Stage I: Tagging above). A non-exhaustive list of supported applications includes genome re-sequencing, histone or protein ChIP-Seq (chromatin immunoprecipitation sequencing) or ATAC-Seq (Assay for Transposase-Accessible Chromatin by sequencing).\n\nA use case of ChIP-Seq for the transcription factor ZFP57 is shown in Figure 1 (data re-analysed from 17). Alignments to an N-masked reference genome were performed for reciprocal crosses between Black 6 and Cast/EiJ mice using Bowtie 2, followed by SNPsplit sorting. This process was able to identify allele-specific binding of ZFP57 to several different imprinting control regions in a parental origin-specific manner, exemplified for the SNRPN locus in Figure 1.\n\nThe binding of ZFB57 is methylation-dependent and can be found exclusively on the maternal allele (genetic background of mother in forward cross: Black 6; mother in reverse cross: Cast). SNP positions were N-masked and used for allele-specific splitting of sequencing reads (shown as horizontal lines in black). Allele 1: Black 6 reference. Allele 2: Cast/EiJ strain (Cast). The area shown depicts the DMR only in part. Data taken from 17 (GEO accession: GSE55382).\n\nIn addition to standard linear alignments with or without small InDels, SNPsplit also handles spliced read alignments containing large gaps (CIGAR operation: N), such as reads spanning exon boundaries in RNA-Seq experiments. Spliced read aligners that have successfully been used for allele-specific alignments in conjunction with SNPsplit include Tophat18, STAR14 (Spliced Transcripts Alignment to a Reference) and HISAT213. To work smoothly together with SNPsplit, HISAT2 and STAR require the user to disable soft-clipping which is performed by default (CIGAR operation: S), and STAR also needs to be instructed to print out the mismatch position (MD:Z:) field. More detailed instructions may be found in the SNPsplit User Guide.\n\nAs a variant of the chromatin conformation capture assay Hi-C is a proximity-ligation based assay which allows the investigation of the three-dimensional structure of the genome by massively parallel sequencing19. This is accomplished by measuring the frequency at which different parts of the genome sequence come into close physical contact. While standard Hi-C cannot discriminate whether an interacting fragment originated from the same or the other allele, allele-specific interaction maps can separate cis-allele from trans-allele interactions, thereby greatly improving the analysis of chromatin dynamics and gene regulation20, 21.\n\nThe Hi-C mode of SNPsplit assumes that the input data is in the Hi-C format produced by the HiCUP pipeline22, i.e. the input BAM files are by definition paired-end and Read 1 and Read 2 follow each other. It discriminates several additional read combinations to distinguish between cis- and trans-allele interactions:\n\n- G1-G1\n\n- G2-G2\n\n- G1-UA\n\n- G2-UA\n\n- G1-G2\n\n- UA-UA\n\nFor mixed allele groups such as G1-G2 there is no need to create the reverse group (G2-G1) since Hi-C interactions have no directionality. Again, read pairs containing at least one conflicting read (tag: CF) are not printed out by default, but this may be optionally enabled.\n\nBisulfite sequencing is a method to interrogate DNA methylation patterns using the chemical properties of sodium bisulfite to convert cytosines to uracil but leaving methylated cytosines largely unaffected.\n\nThe bisulfite mode of SNPsplit assumes that the input data has been processed with the bisulfite alignment tool Bismark23. SNPsplit runs a quick check at the start of a run to see if the file provided appears to be a Bismark file, and sets the appropriate flags for bisulfite and/or paired mode automatically. Paired-end mode requires Read 1 and Read 2 of a pair to follow each other in consecutive lines so the BAM file will be sorted by read name if necessary.\n\nUtilisation of SNP positions and allele assignment of bisulfite treated reads In contrast to the standard mode, C>T SNPs may not always be used for allele-specific sorting in a bisulfite setting since they could either be a genuine SNP or rather reflect the methylation state. Since the majority of known SNPs actually involves C to T transitions (due to spontaneous deamination of methylated CpG dinucleotides), the ability to assign aligned bisulfite treated reads is thus somewhat reduced compared to regular DNA-based alignments. The number of SNP positions that have been skipped because of this bisulfite ambiguity is documented in the report file.\n\nPositions requiring special treatment include all of the following Allele 1/Allele 2 combinations: C/T or T/C for forward strand alignments and G/A or A/G for reverse strand alignments. These positions may however be used to assign opposing strand alignments since they do not involve C to T transitions directly. For that reason, the bisulfite call processing also extracts the bisulfite strand information from the alignments in addition to the basecall at the position involved. For any SNPs involving C positions that are not C to T SNPs both methylation states, i.e. C and T, are allowed to match the C position.\n\nFor SNPs which were masked by Ns in the genome no methylation call will have been performed during the alignment step, i.e. they will receive a ‘.’ (dot) in the methylation call string. This means that SNP positions themselves may be used for allele-sorting but do not participate in calling methylation. While this reduces slightly the number of total methylation calls it effectively eliminates the problem of assigning potentially incorrect methylation states to these positions.\n\nTo demonstrate the effectiveness of sorting bisulfite treated reads we reprocessed publicly available bisulfite sequencing data from reciprocal mouse crosses reported by Xie and colleagues6 (GEO accession: GSE33722). First we generated a dual hybrid genome for 129X1/SvJ (129) (as near-enough relative we used the SNP annotations for strain 129S1/SvImJ) and Cast/EiJ (Cast) mice, and then aligned the data to the N-masked genome using Bismark (v0.16.1, default parameters, read trimming was performed using Trim Galore24 v0.4.1, default parameters). The data was then processed with SNPsplit and all datasets for the F1 forward cross 129 (mother) x Cast (father), and F1 reverse cross Cast (mother) x 129 (father) were merged and analysed using SeqMonk (v.0.33.025) (Figure 2).\n\nThe upstream DMR is methylated exclusively on the paternal allele, while the more downstream DMR is methylated exclusively on the maternal allele in both forward (129 x Cast) and reverse crossed (Cast x 129) hybrid mice. SNP positions were N-masked and used for allele-specific sorting with SNPsplit. Allele 1: 129X1/SvJ reference (129). Allele 2: Cast/EiJ strain (Cast). Red or blue dots in the graph represent calls for methylated or unmethylated cytosines, respectively (CpG context only). The percentage methylation was determined for 2000 bp windows for the region shown using the Bisulfite Methylation Pipeline in Seqmonk25 (default options). Data taken from 6, GEO accession: GSE33722.\n\nWhilst the majority of the genome shows very similar methylation levels on both alleles of the hybrid mice, this approach also allows the detection of allele-specific methylation events. This can be readily spotted at imprinted loci where one parental allele is fully methylated while the other remains completely unmethylated. The Gnas/Nespas locus in the mouse genome shows both a paternally methylated region (more upstream) and maternally methylated region (more downstream) where the allele-specific methylation pattern is maintained in a parent-of-origin dependent manner (Figure 2). This demonstrates that the combination of bisulfite mapping and read sorting by SNPsplit is an effective tool to identify allele-specific methylation in diploid genomes.\n\n\nSummary\n\nAnalysing next-generation sequencing data in an allele-specific fashion holds the potential to uncover regulatory events or mechanisms that would otherwise be obscured in bulk data. SNPsplit is designed to enable researchers to quickly and easily perform allele-specific analysis of their sequencing data as long as the SNP genotypes of the organism in question are known. For hybrid mouse strains covered by the Mouse Genomes Project, SNPsplit offers an easy solution from generating N-masked genomes to allele-specific sorting of reads without requiring the user to possess excessive computational skills. SNPsplit is not tied to any specific application and indeed it has been used already to answer questions for a variety of different data types such as ChIP-Seq, RNA-Seq, Bisulfite-Seq and Hi-C. This gives SNPsplit the unique capability of bringing together allele-specific data including gene-expression, DNA methylation, genomic accessibility or architecture which holds great potential for studying genome regulation.\n\n\nSoftware availability\n\n1. Software available from: http://www.bioinformatics.babraham.ac.uk/projects/SNPsplit/\n\n2. Latest source code: https://github.com/FelixKrueger/SNPsplit\n\n3. Archived source code as at time of publication: https://zenodo.org/record/55477#.V18PoDb93ww28\n\n4. Software license: GNU GPL v3 or later",
"appendix": "Author contributions\n\n\n\nFK designed and wrote SNPsplit and the manuscript, SRA was involved in study design and wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nResearch was supported by the Babraham Institute and the UK Biotechnology and Biological Sciences Research Council (BBSRC).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nSelvaraj S, R Dixon J, Bansal V, et al.: Whole-genome haplotype reconstruction using proximity-ligation and shotgun sequencing. Nat Biotechnol. 2013; 31(12): 1111–1118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRivas MA, Pirinen M, Conrad DF, et al.: Human genomics. Effect of predicted protein-truncating genetic variants on the human transcriptome. Science. 2015; 348(6235): 666–669. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDegner JF, Marioni JC, Pai AA, et al.: Effect of read-mapping biases on detecting allele-specific expression from RNA-sequencing data. Bioinformatics. 2009; 25(24): 3207–3212. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrowley JJ, Zhabotynsky V, Sun W, et al.: Analyses of allele-specific gene expression in highly divergent mouse crosses identifies pervasive allelic imbalance. Nat Genet. 2015; 47(4): 353–360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRozowsky J, Abyzov A, Wang J, et al.: AlleleSeq: analysis of allele-specific expression and binding in a network framework. Mol Syst Biol. 2011; 7(1): 522. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXie W, Barr CL, Kim A, et al.: Base-resolution analyses of sequence and parent-of-origin dependent DNA methylation in the mouse genome. Cell. 2012; 148(4): 816–831. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan de Geijn B, McVicker G, Gilad Y, et al.: WASP: allele-specific software for robust molecular quantitative trait locus discovery. Nat Methods. 2015; 12(11): 1061–1063. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeane TM, Goodstadt L, Danecek P, et al.: Mouse genomic variation and its effect on phenotypes and gene regulation. Nature. 2011; 477(7364): 289–294. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCastel SE, Levy-Moonshine A, Mohammadi P, et al.: Tools and best practices for data processing in allelic expression analysis. Genome Biol. 2015; 16: 195. PubMed Abstract | Publisher Full Text | Free Full Text\n\nServant N, Varoquaux N, Lajoie BR, et al.: HiC-Pro: an optimized and flexible pipeline for Hi-C data processing. Genome Biol. 2015; 16: 259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangmead B, Salzberg SL: Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012; 9(4): 357–359. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H, Durbin R: Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics. 2010; 26(5): 589–595. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim D, Langmead B, Salzberg SL: HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015; 12(4): 357–360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–2079. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQC Fail. 2016. Reference Source\n\nStrogantsev R, Krueger F, Yamazawa K, et al.: Allele-specific binding of ZFP57 in the epigenetic regulation of imprinted and non-imprinted monoallelic expression. Genome Biol. 2015; 16(1): 112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim D, Pertea G, Trapnell C, et al.: TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013; 14(4): R36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLieberman-Aiden E, van Berkum NL, Williams L, et al.: Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science. 2009; 326(5950): 289–293. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDixon JR, Jung I, Selvaraj S, et al.: Chromatin architecture reorganization during stem cell differentiation. Nature. 2015; 518(7539): 331–336. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRao SS, Huntley MH, Durand NC, et al.: A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014; 159(7): 1665–1680. PubMed Abstract | Publisher Full Text\n\nWingett S, Ewels P, Furlan-Magaril M, et al.: HiCUP: pipeline for mapping and processing Hi-C data [version 1; referees: 2 approved, 1 approved with reservations]. F1000Res. 2015; 4: 1310. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrueger F, Andrews SR: Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics. 2011; 27(11): 1571–1572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrim Galore. Reference Source\n\nSeqMonk. Reference Source\n\nMifsud B, Tavares-Cadete F, Young AN, et al.: Mapping long-range promoter contacts in human cells with high-resolution capture Hi-C. Nat Genet. 2015; 47(6): 598–606. PubMed Abstract | Publisher Full Text\n\nSchoenfelder S, Furlan-Magaril M, Mifsud B, et al.: The pluripotent regulatory circuitry connecting promoters to their long-range interacting elements. Genome Res. 2015; 25(4): 582–597. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrueger F: SNPsplit: Allele-specific splitting of alignments between genomes with known SNP genotypes. Zenodo. 2014. Publisher Full Text"
}
|
[
{
"id": "14558",
"date": "01 Jul 2016",
"name": "Andrew Keniry",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nKrueger and Andrews report SNPsplit, a tool for sorting mapped next generation sequencing reads into separate files depending on the allelic origin of the read. The process requires genetic heterogeneity between the alleles such that the reads can be assigned to a particular allele based on known SNP positions. Such a tool allows for allele specific analysis and becomes useful for studies on phenomena such as genomic imprinting, allelic imbalance and X-chromosome inactivation. SNPsplit improves on current methods for analysis of allele specific sequencing reads by providing built in tools for the analysis of bisulfite and HiC data, which are otherwise more complicated to analyse. A tool for creating an N-masked genome is also provided, which overcomes mapping bias towards the reference genome.\n\nTo date I have used SNPsplit to process data from RNA-seq, bisulfite-seq, transcription factor ChIP-seq and histone ChIP-seq. I have not tested the built in N-masked genome creator. SNPsplit has proven to be easy to use and very stable in my run environment. Typically, a sample is processed in approximately 3 hours depending on read depth. As far as I can tell, by assessing known imprinted genes and the silent female X chromosome, SNPsplit does an excellent job of assigning reads to the correct genome, with known phenomena appearing as expected. The built in option for analysis of bisulfite data works very well on reads mapped with the bismark program.\n\nI’ve found SNPsplit to work very well for all the data types I have used it for: RNA-seq, bisulfite-seq, transcription factor ChIP-seq and histone ChIP-seq. It should be noted however that due to the requirement for a SNP to be present in the read, assignable reads from narrow transcription factor ChIP-seq peaks can be sparse and some peaks may not assignable at all. This is an unavoidable limitation, and the authors show successful assignment of transcription factor ChIP-seq reads in Figure 2, however this is perhaps something researchers should consider. Reads deriving from broad histone ChIP-seq peaks suffer no such limitation.\n\nI have no issues with the performance of SNPsplit, and neither the accuracy of how the tool is presented in the paper. Perhaps the authors could include some details that will aid researchers in experimental design? For example, for a typical SNP density what percentage of reads would be assigned to each genome, unassigned and conflicting? This would help researchers in estimating required read depths for their particular question. Could the authors also explain how vcf files are filtered for high confidence SNPs when preparing the N-masked genome? A discussion of post processing techniques for the removal of incorrectly annotated SNPs would also be beneficial. Perhaps the benefit of longer read length for assigning reads to a particular genome could also be mentioned?\n\nIn summary, SNPsplit will prove to be a very useful tool for the analysis of epigenomic data from next generation sequencing experiments.",
"responses": [
{
"c_id": "2093",
"date": "27 Jul 2016",
"name": "Felix Krueger",
"role": "Author Response",
"response": "We would like to thank the reviewer for their kind and positive feedback. Please find our specific replies below. Could the authors also explain how vcf files are filtered for high confidence SNPs when preparing the N-masked genome? We have expanded the SNPsplit User Guide considerably to now include a detailed section on how the SNP filtering is accomplished and which of the parameters are required for the process to work with other VCF files. We hope that this will make the process easier to understand and allow adapting the procedure for other genomes as well. Perhaps the authors could include some details that will aid researchers in experimental design? For example, for a typical SNP density what percentage of reads would be assigned to each genome, unassigned and conflicting? This would help researchers in estimating required read depths for their particular question. A discussion of post processing techniques for the removal of incorrectly annotated SNPs would also be beneficial. Perhaps the benefit of longer read length for assigning reads to a particular genome could also be mentioned? It would be undoubtedly helpful to provide some more guidance about experimental design, however we are not sure that this should be added in the context of software manuscript. As a general guideline the percentage of reads that can be assigned allele specifically increases proportionally with the number of heterozygous SNPs present between two strains, and increasing the read length also increases the chances to hit a SNP. Furthermore paired-end data can be assigned with a much rate than single-end reads. The SNPsplit User Guide contains a few example reports to help users to get an idea about typical values. We also feel that post-processing techniques probably warrant more discussion than just being briefly mentioned here, a good paper to read to get started is also cited in the manuscript (Castel et al.)."
}
]
},
{
"id": "14557",
"date": "01 Jul 2016",
"name": "Nicolas Servant",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript by Krueger and collaborators describes SNPsplit, an alignment sorter for allele specific analysis. SNPsplit is designed to work on N-masked alignment and provides additional utilities to generate the appropriate reference.\nThe main interest of SNPsplit is its ability to operate across different experiment types and alignment software. It therefore allows to analyse and to integrate in an allele specific and unbiased manner heterogeneous dataset.\nThe manuscript is well written, and is divided in two main parts. The first part presents the software and its implementation and the second part, presents user cases.\nOther than my few comments below, I am happy with the manuscript and want to note that I have successfully downloaded and used SNPsplit in the context of several projects.\nThe way the N-masked genome is generated based on SNPs information is not that easy to understand for non expert users. I would recommend a figure to help in understanding this point.\nIn the context of Dual-Hybrid strain, the interest of first generating the strain 1 (S1) and the strain 2 (S2) genomes is not clear to me.\nIf I'm correct, the final masked genome will be generated from the S1 fasta reference only (unmasked). If we use a simple example with a SNP which is Bl6=A, S1=T, S2=T, using this strategy will avoid a mismatch at this position, compared to the strategy of masking only heterozygous SNPs between S1/S2 on the reference (Bl6) genome. But what is the interest of generating the S2 reference genome (step 4) ? Using S1 or S2 should be enough?\n\nHow paired-end sequencing is managed in practice ? The authors present an application with Hi-C data but is it the same for any NGS application?\n\nThe authors present SNPsplit in the context of the Mouse Genome Project. I'm wondering how difficult it would be to transpose it to any genome and organism as long as the genotype is known. A few words about that in the manuscript would be interesting.\n\nIn the Introduction, the authors mentioned the WASP software[7]. To avoid any misunderstanding with the parental genome strategy, I would suggest something like; “A more recent improvement aims to reduce mapping biases by first aligning reads on the reference genome, then realigning reads that overlap SNP positions in all possible allele combinations and keeping only reads that align to the same position regardless their genotypes - this reduces bias, but is computationally complex”\n\nAt the end of the introduction, the authors mentioned that “SNPsplit is not tied to any particular aligner” which is not exactly true as the authors explain later that “SNPsplit can be used as long as the CIGAR operation requirements are met”\n\nTypo:\n“The determination of overlaps is geared” in Stage I: Tagging Reference 26 and 27 doesn't seem to be used in the text",
"responses": [
{
"c_id": "2092",
"date": "27 Jul 2016",
"name": "Felix Krueger",
"role": "Author Response",
"response": "Many thanks for the constructive review and the thoughtful suggestions on how to improve the manuscript. Please find our point-by-point responses below. The way the N-masked genome is generated based on SNPs information is not that easy to understand for non expert users. I would recommend a figure to help in understanding this point. We have now added a new Figure 1 that aims to explain the process of generating single or dual hybrid N-masked genomes in more detail. We feel that this new figure makes the whole process substantially easier to understand, many thanks for suggestion! In the context of Dual-Hybrid strain, the interest of first generating the strain 1 (S1) and the strain 2 (S2) genomes is not clear to me. If I'm correct, the final masked genome will be generated from the S1 fasta reference only (unmasked). If we use a simple example with a SNP which is Bl6=A, S1=T, S2=T, using this strategy will avoid a mismatch at this position, compared to the strategy of masking only heterozygous SNPs between S1/S2 on the reference (Bl6) genome. But what is the interest of generating the S2 reference genome (step 4) ? Using S1 or S2 should be enough? Technically it would be sufficient to use the Strain 1 unmasked genome as new reference and then use the SNP annotations of both strains to compute a new Strain1/Strain2 SNP annotation file (see also the new Figure 1). Since writing out a new genome only takes a few seconds and it might be of potential use for other projects we do also write out versions of Strain 2, it can always be removed later if desired. How paired-end sequencing is managed in practice? The authors present an application with Hi-C data but is it the same for any NGS application? We do mention in the manuscript that the paired-end mode uses both ends for the genome-specific assignments even though it is not as complicated as the Hi-C mode. The SNPsplit User Guide already contains more detailed information about paired-end file handling and also provided paired-end sorting reports to illustrate this, so we would kindly refer the user to the manual rather than adding an extra section here. The authors present SNPsplit in the context of the Mouse Genome Project. I'm wondering how difficult it would be to transpose it to any genome and organism as long as the genotype is known. A few words about that in the manuscript would be interesting. Indeed, as long as the genotypes are well defined and the list of SNP positions is known the genome preparation should also work well with any other genome (see also the reply to Prasoon Agarwal’s comment). To enable this process we have added a new section to the SNPsplit User Guide that explains the SNP filtering and processing in more detail, and we have added a short paragraph about this to the manuscript. In the Introduction, the authors mentioned the WASP software[7]. To avoid any misunderstanding with the parental genome strategy, I would suggest something like; “A more recent improvement aims to reduce mapping biases by first aligning reads on the reference genome, then realigning reads that overlap SNP positions in all possible allele combinations and keeping only reads that align to the same position regardless their genotypes - this reduces bias, but is computationally complex” We have changed this sentence to make it clearer. At the end of the introduction, the authors mentioned that “SNPsplit is not tied to any particular aligner” which is not exactly true as the authors explain later that “SNPsplit can be used as long as the CIGAR operation requirements are met” At least in its generic mode SNPsplit is not tied to any specific aligner, but this does not necessarily mean that every single option of any aligner will be supported. To this end, SNPsplit was initially designed to work with Bowtie2 and Tophat, but we have later also seen other software such as STAR or HISAT2 work fine as well. Aligner-specific comments can be found in the SNPsplit User guide. Reference 26 and 27 doesn't seem to be used in the text Thanks for spotting this, we have now removed these references from the bibliography."
}
]
},
{
"id": "14556",
"date": "08 Jul 2016",
"name": "Prasoon Agarwal",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article by Krueger F et al. describes in detail a new software SNPsplit which is capable of sorting reads or read–pairs in the allele specific modus that covers know single nucleotide polymorphic (SNP), or single nucleotide variation (SNV) locations. For sorting the reads the inputs to the software are a ‘N’ masked genome, created using known SNP or SNV locations, to which the reads are aligned using any available aligner and a VCF file containing the positions of all SNPs. The authors claim that this is the only user-friendly solution available for the allele-specific splitting of sequencing reads aligned to N-masked genomes. Overall the manuscript is very well written. However, I have some minor queries regarding the software performance:\nMasking known SNP positions in the genome sequence eliminated the reference bias but, in case of heterozygous SNPs there could be a chance of having significant bias toward higher mapping rates of the allele in the reference sequence, is there any provision in the software to remove this noise and bias from mapped reads? This kind of bias can lead to false signal of allelic imbalance.\n\nIt is stated that the software can construct the N-masked genomes, so is it restricted to mouse alone or can be extended in case of humans or other species?\n\nCan SNPsplit be used to split reads for indels and deletions?\nI have personally used the software for ATAC-seq data from patients and it works perfectly fine. I have not used the genome builder module of the software. The data looks perfect in the UCSC genome browser for the ‘Allele 1’ or ‘Allele 2’ and the unassigned. Overall I felt it is a very user friendly software available.",
"responses": [
{
"c_id": "2091",
"date": "27 Jul 2016",
"name": "Felix Krueger",
"role": "Author Response",
"response": "We would like to thank the reviewer for their kind and approving comments about SNPsplit. Some specific comments may be found below. Masking known SNP positions in the genome sequence eliminated the reference bias but, in case of heterozygous SNPs there could be a chance of having significant bias toward higher mapping rates of the allele in the reference sequence, is there any provision in the software to remove this noise and bias from mapped reads? This kind of bias can lead to false signal of allelic imbalance. SNPsplit is really designed for mapping against genomes with clean parental genotypes; this is also why the genome preparation step is very strict and only uses high quality homozygous parental SNPs (see the new section on this in the SNPsplit manual). If the SNPs are heterozygous already in one or even both of the parental strains we propose that these positions could be masked by Ns for the mapping step, but not be used at all for the allele-specific read assignment as such. For the allele-specific splitting, i.e. the file supplied to SNPsplit with --snp_file, only homozygous high quality SNP positions should be used. It is stated that the software can construct the N-masked genomes, so is it restricted to mouse alone or can be extended in case of humans or other species? The SNPsplit approach does in theory work on any N-masked genome irrespective of the species. The preparation is currently optimised to work the VCF file provided by the Mouse Genomes Project but as long as the VCF file conforms to the same standards it should also work for other genomes. If the VCF file looks different, e.g. you got the file from a collaborator and it came without header lines or format description one would have to adapt the relevant section(s) in the preparation script. We have now added a section to the SNPsplit User Guide that explains in more detail the basis of SNP filtering and which parameters are needed to run properly. We hope that this will help generating N-masked genomes when the SNP data looks different than the file provided by Mouse Genomes Project. Can SNPsplit be used to split reads for indels and deletions? SNPsplit supports the processing of reads that contain indels, but indels themselves are not used to sort reads to different genomes at the current time. The reason for this is mainly that indel annotations are often more tricky to come by and alignments over indels (e.g. at the ends of reads) are notoriously more difficult. There are currently no plans to extend the functionality to include indels, but we might look into this in the future if the demand arises."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1479
|
https://f1000research.com/articles/5-1587/v1
|
06 Jul 16
|
{
"type": "Research Note",
"title": "A multi-site cutting device implements efficiently the divide-and-conquer strategy in tumor sampling",
"authors": [
"Jose I. Lopez",
"Jesus M. Cortes"
],
"abstract": "We recently showed that in order to detect intra-tumor heterogeneity a Divide-and-Conquer (DAC) strategy of tumor sampling outperforms current routine protocols. This paper is a continuation of this work, but here we focus on DAC implementation in the Pathology Laboratory. In particular, we describe a new simple method that makes use of a cutting grid device and is applied to clear cell renal cell carcinomas for DAC implementation. This method assures a thorough sampling of large surgical specimens, facilitates the demonstration of intratumor heterogeneity, and saves time to pathologists in the daily practice. The method involves the following steps: 1. Thin slicing of the tumor (by hand or machine), 2. Application of a cutting grid to the slices (e.g., a French fry cutter), resulting in multiple tissue cubes with fixed position within the slice, 3. Selection of tissue cubes for analysis, and finally, 4. Inclusion of selected cubes into a cassette for histological processing (with about eight tissue fragments within each cassette). Thus, using our approach in a 10 cm in-diameter-tumor we generate 80 tumor tissue fragments placed in 10 cassettes and, notably, in a tenth of time. Eighty samples obtained across all the regions of the tumor will assure a much higher performance in detecting intratumor heterogeneity, as proved recently with synthetic data.",
"keywords": [
"Tumor sampling",
"cutting grid",
"divide and conquer",
"clear cell renal cell carcinoma",
"intratumor heterogeneity",
"pathology routine"
],
"content": "Introduction\n\nIn the light of current findings provided by numerous sequencing tools, it is known that practically all human neoplasms display some degree of intratumor heterogeneity (ITH)1. Characteristically, ITH is not uniformly distributed along the tumor; instead, it shows a regional distribution following a stochastic pattern, the final result being unique, unpredictable, and dynamically varying along the time2. The correct identification of ITH is mandatory now that targeted therapies are offering promising results to patients3, but pathologists - the specialists in charge of tumor selection for analysis - seem to have not given, so far, an appropriate answer to this issue.\n\nWe have recently proposed a reliable, affordable and time-saving solution to this problem4. The goal is twofold; to improve ITH detection and to perform ITH at affordable laboratory costs. This simple solution is based on the divide-and-conquer algorithm (DAC). Noteworthy, tumor sampling following DAC outperforms the routine protocol sampling for identifying ITH and, it does it, at a similar cost4. However, pathologists could consider that DAC is a time-consuming method when grossing, which might make it difficult to introduce it in routine practice. In this brief report, we describe a simple procedure to overcome this problem.\n\n\nMethod and results\n\nThe so-called DAC algorithm5 is based on recursively breaking down a problem in smaller parts (divide) until these are simple enough to be solved directly (conquer). Then, partial solutions are combined to solve the original problem. DAC strategies have been largely applied in science to solve complex problems, including several challenging issues in biomedical areas. As early as in 1967, DAC helped clinicians to correlate hypoglycemia with infantile convulsions6. In addition, DAC has been useful in cell biology and oncology, for instance in selecting the appropriate cells for biological experiments7 and, more recently, in helping to decipher breast cancer heterogeneity8.\n\nHere DAC is applied to clear cell renal cell carcinomas (ccRCCs), since these tumors are frequently large and, for this reason, impossible to be totally sampled. Any other large tumor, however, can benefit from this method. The DAC strategy (Figure 1) requires the pathologist to select, instead of a few large fragments, a substantial number of small ones widely distributed along the entire tumor. However, pathologists under a daily routine pressure can perceive this method as laborious and time-consuming.\n\nA simple device consisting of a cutting grid (here, a potato cutter) will overcome this inconvenience. When applied directly to the whole tumor surface previously thin-sliced, the grid will cut it into small cubes in one shot (Figure 2). Next, the pathologist’s decision will consist simply in selecting the cubes that will be processed for analysis as previously reported4. The method can be applied (and has been tested) to both fresh and formalin-fixed tissue, saves time, and assures a uniform sampling distribution along the tumor. The objective for improving efficiency of targeted therapies is the discovery of the complete ITH spectrum, and not its exact location. Thus the selected cubes included in the cassettes (six to eight cubes per cassette) will provide much more thorough information of the tumor, both under the microscope as well as at the molecular level.\n\n\nDiscussion\n\nThe use of the DAC method to help sampling strategies is not new. Indeed some authors have applied the algorithm in particle physics to improve the diffusion sampling in generalized ensemble simulations9. This approach, or any other with scientific basis, has not been implemented for tissue selection in Pathology laboratories so far, since the pathologists did not consider tumor sampling a complex problem in the pre-molecular era.\n\nAn experience-based reasoning says that this option will save pathologist’s time when handling large tumors, in a manner which is inexpensive and reliable at the same time. In combination with the changes proposed for the technician training in our previous report4, this new alternative will make the pathologists’ routine much faster and robust providing an integrated solution to fulfill basic researchers’ expectations10. If the DAC strategy is adopted as a suitable method to increase the amount of information given to oncologists, pathologist's routine will move from the classic big-fragments-into-the-cassette routine to a sort of rudimentary tissue microarray building, as recently proposed4.\n\nFigures are demonstrative. For instance, the DAC strategy applied to a ccRCC of 10 cm in diameter - a quite common situation in routine pathology - will generate approximately 80 small samples (of about 4–5 mm in size) that would be included in 10 cassettes for a thorough tumor examination. Importantly to remark, the same 10 cm in diameter tumor would need also 10 cassettes for the analysis, with one tumor sample per cassette, in the case of routine sampling protocols11.\n\nDepending on the pathologist’s skills, the time to collect 80 small samples in the grossing room is variable, but in any case, long. For this reason, any successful alternative must necessarily overcome this hurdle. A feasible choice would be an electric bacon slicer, but a long bladed knife will also work. To note, slicing electric machines are being increasingly used in pathology for handling radical prostatectomies12 and other surgical specimens13, and they are the first step in the whole-mounting processing for tumor mapping. In this case, the obtained ccRCC slices can be quickly cut in one shot by pressing on the entire tumor surface with a cutting grid. The procedure will generate many cubes ready to be included within a cassette. If we assume that tumor sampling following a DAC strategy is appropriate for improving ITH detection, the use of a cutting grid will shorten significantly the total process ensuring a uniform and widespread selection of the samples. A straightforward estimation with some practical cases indicates that the time for obtaining 80 samples with this method is reduced to a tenth as we cut 10 small tumor pieces at the same time.\n\n\nConclusions\n\nThe present paper describes a new method for tumor sampling in routine pathology inspired by the DAC algorithm4. Once DAC has been proved to be efficient for ITH detection, we expect that the use of a cutting grid will make affordable its widespread application. Objectives are twofold: ITH detection improvement and time optimization (cost) in Pathology laboratories.",
"appendix": "Author contributions\n\n\n\nJIL and JMC identified the problem and gave a realistic solution. JIL and JMC wrote this note.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nJMC is funded by Ikerbasque: The Basque Foundation for Science.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nGay L, Baker AM, Graham TA: Tumour Cell Heterogeneity [version 1; referees: 5 approved]. F1000Res. 2016; 5: pii: F1000 Faculty Rev-238. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGerlinger M, Rowan AJ, Horswell S, et al.: Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med. 2012; 366(10): 883–892. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiley C, de Bruin EC, McGranahan N, et al.: Deciphering intratumor heterogeneity and temporal acquisition of driver events to refine precision medicine. Genome Biol. 2014; 15(8): 453. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLópez JI, Cortés JM: A divide-and-conquer strategy in tumor sampling enhances detection of intratumor heterogeneity in routine pathology: A modeling approach in clear cell renal cell carcinoma [version 1; referees: 4 approved]. F1000Res. 2016; 5: 385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCormen TH, Leiserson CE, Rivest RL, et al.: Introduction to Algorithms. 2nd Edition, MIT Press, 2001. Reference Source\n\nDivide and conquer. JAMA. 1967; 202(13): 1144. PubMed Abstract | Publisher Full Text\n\nEisenstein M: Cell sorting: Divide and conquer. Nature. 2006; 441(7097): 1179–1185. PubMed Abstract | Publisher Full Text\n\nKristensen VN: Divide and conquer: the genetic basis of molecular subclassification of breast cancer. EMBO Mol Med. 2011; 3(4): 183–185. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMin D, Yang W: A divide-and-conquer strategy to improve diffusion sampling in generalized ensemble simulations. J Chem Phys. 2008; 128(9): 094106. PubMed Abstract | Publisher Full Text\n\nSoultati A, Stares M, Swanton C, et al.: How should clinicians address intratumour heterogeneity in clear cell renal cell carcinoma? Curr Opin Urol. 2015; 25(5): 358–366. PubMed Abstract | Publisher Full Text\n\nTrpkov K, Grignon DJ, Bonsib SM, et al.: Handling and staging of renal cell carcinoma: the International Society of Urological Pathology Consensus (ISUP) conference recommendations. Am J Surg Pathol. 2013; 37(10): 1505–1517. PubMed Abstract | Publisher Full Text\n\nEgevad L: Handling of radical prostatectomy specimens. Histopathology. 2012; 60(1): 118–124. PubMed Abstract | Publisher Full Text\n\nHelliwell TR: ACP Best Practice No 157. Guidelines for the laboratory handling of laryngectomy specimens. J Clin Pathol. 2000; 53(3): 171–176. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "14803",
"date": "07 Jul 2016",
"name": "Thomas V. Colby",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nA simple and practical proposal for tumor sampling tissue. One suggestion: give some guidance as to which among multiple tissue cubes would be put in the cassettes.",
"responses": []
},
{
"id": "14804",
"date": "08 Jul 2016",
"name": "Christopher DM Fletcher",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is valuable and innovative work, which provides a simple and practical approach to more efficient sampling of larger human tumors, with the goal of better determining tumor heterogeneity.\nThere are some omissions in the description of the technique, which would make adoption of the technique easier for others. Specifically, the authors should state clearly the recommended thickness of the whole tumor slice, whether using a long-bladed knife or slicing machine. Clearly the slice thickness must fit in the tissue processing cassette - so probably less than 3 mm - do the authors think that 3 mm slices of a large mass will generally remain intact/not fall apart before being cut with the grid. The second, perhaps more important, point is the authors need to state clearly how much of a given tumor they believe should be sampled. In their 10 cm example, it seems like they would submit virtually the whole tumor but this is not clear. How much would they submit from a 20 cm tumor? Conversely, how big does a tumor need to be before this type of sampling technique is advised? Presumably smaller tumors (e.g. 5 cm) could be sampled in the conventional/routine fashion with larger pieces? Clarification of these points will enhance the value of the authors' proposal.",
"responses": []
},
{
"id": "15138",
"date": "21 Jul 2016",
"name": "Manuel Areias Sobrinho Simoes",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article “A multi-site cutting device implements efficiently the divide-and-conquer strategy in tumor sampling” represents an intelligent and practical evolution of the previous article of the same authors on the challenging issues of intratumor heterogeneity and tumor sampling.\nTumor heterogeneity is one of the most important features of any approach to oncobiology and/or oncology. Together with other factors with which it is related – topography and time (evolution) – tumor heterogeneity is a key issue of cancer diagnosis and cancer treatment.\nLópez and Cortes addressed this issue adequately in their previous article demonstrating the importance of finding ways of evaluating tumor heterogeneity and topography via a clever and innovative strategy – Divide-and conquer (DAC) strategy – of tumor sampling.\nI must confess I liked very much the previous article but I did not know how much effort one would need to implement a successful DAC strategy. It was therefore an agreeable surprise to see the continuation of the work having the real life of a Pathology Laboratory as the objective.\nI think the authors describe reasonably well the method they have “invented” and I concur with them it looks a simple procedure that assures a rather complete sampling of large surgical specimens in a relatively short period of time. However, for the sake of utility, I think the authors should provide a more detailed description of the “technicalities” of the procedure.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1587
|
https://f1000research.com/articles/5-766/v1
|
27 Apr 16
|
{
"type": "Opinion Article",
"title": "“The molecule’s the thing:” the promise of molecular modeling and dynamic simulations in aiding the prioritization and interpretation of genomic testing results",
"authors": [
"Gavin R. Oliver",
"Michael T. Zimmermann",
"Eric W. Klee",
"Raul A. Urrutia",
"Michael T. Zimmermann",
"Eric W. Klee",
"Raul A. Urrutia"
],
"abstract": "Clinical genomics is now a reality and lies at the heart of individualized medicine efforts. The success of these approaches is evidenced by the increasing volume of publications that report causal links between genomic variants and disease. In spite of early success, clinical genomics currently faces significant challenges in establishing the relevance of the majority of variants identified by next generation sequencing tests. Indeed, the majority of mutations identified are harbored by proteins whose functions remain elusive. Herein we describe the current scenario in genomic testing and in particular the burden of variants of unknown significance (VUSs). We highlight a role for molecular modeling and molecular dynamic simulations as tools that can significantly increase the yield of information to aid in the evaluation of pathogenicity. Though the application of these methodologies to the interpretation of variants identified by genomic testing is not yet widespread, we predict that an increase in their use will significantly benefit the mission of clinical genomics for individualized medicine.",
"keywords": [
"Individualized Medicine",
"Clinical Genomics",
"Diagnostic Odyssey",
"Oncology",
"Molecular Modeling",
"Molecular Dynamics",
"Variants of Unknown Significance"
],
"content": "\n\nIn under a decade, sequencing of a human genome moved from a three billion dollar, multi institutional effort, to a common research assay. Now, 16 years after completion of the draft sequence, genomic testing is an increasingly prevalent component of clinical testing. Oncology, hematology and the diagnosis of rare genetic disorders (diagnostic odyssey) have particularly benefited from the increased genetic testing resolution afforded by modern sequencing technologies. The commercialization of clinical assays to profile patients’ somatic or germline genomes is creating the potential for higher resolution diagnoses and individualized treatment options. Stories of the resounding success of such efforts have been widely and justifiably publicized, and have now captured imaginations well beyond the laboratory or clinic; even so far as the White House and US Congress. The President’s Precision Medicine Initiative1 has now been signed into existence and guarantees expanded exploration within this burgeoning area of health sciences research.\n\nFor the researchers and clinicians on the ground of this nascent field, enthusiasm is high but it is tempered with frustration due to the as-yet high rate of cases for which genomic testing remains insufficiently informative (around 75% for diagnostic odyssey)2,3. While the explosion of genomics research has increased our understanding of the human genome and the wide-ranging functions it encodes, most regions of the genome remain uncharacterized and poorly understood. The majority of genomic sequence variants detected in clinical genomic testing go unreported or are classified as variants of unknown significance (VUSs) due to the lack of understanding of their potential to affect normal physiological function.\n\nA recurring scene for many of the clinicians, genetic counselors and researchers who are adopting and exploring this new arena is the convening of large, multidisciplinary teams of clinical and research staff to pore over lengthy lists of genomic sequence variants, exchange professional opinion and debate next steps. However, in the end, these involved efforts often fail to establish variants with highly confident causal or mechanistic relationships with the disease phenotype. In oncology, the knowledge gained for treatment selection is often compelling, but with little prior evidence, most are difficult to actualize. For example, identification of a novel missense mutation in the functional domain of a known, druggable oncogene might logically appear to be a therapeutic target, but no information may exist linking the mutation to drug efficacy. It is this disconnect between novel genotypes identified and their link to the patient’s phenotypes which thwarts our ability to further improve clinical decision making. Thus, there is a significant need to overcome this critical challenge to expand the success of genomic testing in individualized medicine.\n\nWet-lab assays designed to ascertain functional relevance of specific mutations are a highly desired solution, but they remain cumbersome, time-consuming and cost prohibitive in the great majority of cases. They are therefore misaligned with the high throughput nature of modern genomics. Meanwhile, computational methods of predicting the pathogenicity of genomic sequence alterations exist and are both amenable to high-throughput predictions and widely applied in the field4. These algorithms utilize varied information including evolutionary conservation, genomic position or basic protein-level structural information to assign probabilistic scores or categorical predictions of pathogenicity. Their accuracy varies widely, with alternative tools often producing conflicting predictions for the same variant. Furthermore, the predictions often lack contextualization in terms of biological effect. Algorithmic categorization of a variant as pathogenic offers little insight into the nature of its phenotypic effect and less indication still of whether it is relevant clinically or how to act upon it. Many researchers continue to work to improve these methods, but the burden of VUS interpretation persists and the need for a means to address this burden and facilitate clinical interpretation is widely felt within the field.\n\nThe aforementioned computational tools largely used leverage genome centric information. While there is much proven value to these data, the biological reality comprises many additional layers of complexity. Mutations affect atomic-level biophysical changes that may alter the structure and biochemical function of the genome's protein products. While we detect variants in DNA, we typically interpret their effect by inferring or confirming their deleterious effect on encoded proteins (Figure 1). Even the effects of regulatory variants are typically interpreted as altering the probability of a protein to be expressed or spliced into a given functional form. Thus ideally, any method used to assess pathogenicity of genomic variants should possess the ability to look beyond genomic annotation to a functional context, at an atomic resolution. This increased resolution is likely to yield information that can add context to mutations, better identify the mechanism of pathogenicity, and combine with existing knowledge to facilitate in clinical decision making.\n\nHIV-1 protease has become a model system because of its disease relevance, the availability of mutational and drug binding data, and for its tractable size and molecular stability. The protein’s function is to cleave HIV peptides into the functional proteins of the infectious HIV virion. A) Ligand binding residues are spatially separated. The functional protein dimer is shown with a pharmacologic inhibitor bound to the active site. Residues that are within 3.5Å of the inhibitor are highlighted in tan. The primary sequence is colored identically to the three-dimensional structure to indicate relative positioning of residues. It is apparent that the ligand binding portion of the protease consists of residues that are non-adjacent within the primary sequence, illustrating an advantage of modelling over linear sequence analysis. B) Drug resistance mutations tend to occur in residues within the active site. Many mutations have been characterized that are associated with resistance to inhibitor drugs. While the sites of these mutations are also disjoint in sequence, nearly all of them fall into the same set of drug binding residues, indicating how modeling can enable prediction of a mutation’s effect. C) Structural effects of mutations beyond the active site. Drug-resistance mutations in non-ligand-binding residues have been shown, using computational experiments, to impact the flexibility of the protein and therefore alter drug binding. Computational modeling has characterized the flexibility of the protein in multiple mutated states, illustrating the potential to predict the functional effect of mutations beyond an active site. D) Dynamics of unbound ligand. Computational studies have the advantage of being able to simulate conditions that are difficult to assay experimentally, such as the dynamics of ligand-free forms in atomic detail.\n\nMotivated by the conditions described, we have begun to apply molecular modeling and dynamic simulation techniques in the interpretation of genomic variants identified by next generation sequencing. These methodologies abandon the practice of regarding mutations as occurring in linear strings of nucleotides or amino acids and instead offer a three dimensional, dynamic view, at an atomic resolution. They involve the computational generation, optimization and verification of a protein model, often based on homology to an experimentally determined protein structure5,6. Ab initio modeling is also possible, albeit with lower confidence. The model itself contains information about the linear amino acid sequence of the protein, along with the relative spatial coordinates of its atoms. Precise mathematical and biophysical parameters in the form of a force field are applied to the model to calculate energetic characteristics of the system. The methods are often mature and under reasonable conditions can be expected to produce a model with error comparable to that of a typical structure solved by nuclear magnetic resonance spectroscopy.\n\nMolecular modeling allows us to visualize the manner in which proteins are folded to create a functional structure and to accurately simulate how this is disrupted by mutational events. Because we can visualize atomic bonds, mutations which disrupt inter-molecular interactions - for example between an enzyme and its substrate - can also be modeled. To illustrate their function, many enzymes are analogized to hand tools. One example likens proteases and scissors; if a variant inhibits the closing motion of the scissors about their fulcrum, the blades cannot function; if a variant blocks entry of a material between the blades, then that material can no longer be cut. The reality, of course, is more complex. Proteins are flexible polymers that only achieve mechanistic accuracy by folding into complex and specific three dimensional conformations that restrain or focus thermal fluctuations towards collective motions that are typically part of the mechanism itself. Regions of the properly folded structure’s surface that interact with other molecules, either for chemical modification (e.g. phosphorylation) or structural contacts (e.g. α/β tubulin assembly), can also be critical for function and modified by variants. In addition to the native shape of a protein, the ability of the linear amino acid polymer to achieve that shape is critical. Protein folding often occurs through progressive assembly of local structural elements or intermediates. If an intermediate is either stabilized or destabilized by a variant, the ability of the protein to achieve the native fold could be altered.\n\nComputational biophysics and biochemistry aim to understand molecular function in a dynamic manner at an atomic resolution differing from their wet-lab counterparts methodologically but sharing the same goals. The most obvious advantages of computational approaches include the potential to test hypotheses in silico that would be difficult, costly, or intractable in the lab. Similar to laboratory experiments, computational calculations and simulations are most interpretable when they are well designed, test a specific hypothesis, contain positive and negative controls, connect the data generated to the pre-existent knowledge in the field, and allow drawing further functional inferences. When all these conditions are met, the three dimensional and dynamic representation of the modeled mutations may add a significant value to the interpretation of a genomic test’s finding.\n\nOf course, molecular modeling methods have limitations. The generation of a reliable model is often dependent on the pre-existence of experimentally determined homologous structures and even where these exist, they may provide only partial information7. A model is by nature an approximation of reality. Nonetheless, current techniques have achieved suitable accuracies such that they are frequently used8–11 in applications including drug design, virtual screening, protein engineering and site-directed mutagenesis. With this in mind, their relatively slow uptake in the clinical setting is somewhat surprising. This fact may simply reflect the relative nascence of clinical genomics and the tendency of specialists to seek out fields in which their specialty is already known and accepted.\n\nWe encourage those working within or in proximity to the clinical genomics setting to engage in or promote increased exploration of these methods in their work. Our initial experiences of applying such techniques at the clinical-research boundary of genomics-driven oncology, hematology and diagnostic odyssey have been encouraging and have begun to inform decision-making. We are observing clear initial benefits in regard to variant prioritization, interpretation and validation, with several initial publications in preparation to highlight the value obtained. Of course, not everyone will possess the necessary scientific knowledge or technical skills to deploy these methods directly, but we propose that inter or intra-institutional collaborations may enable those lacking the requisite expertise to identify and access appropriate resources. Furthermore, several freely available online solutions exist12,13 and enable a researcher or clinician to experiment with core modeling methods in the absence of extensive technical expertise. Such independent or collaborative exploration will open the door to deeper understanding of biological mutations and have the potential to inform clinical thinking.\n\nIn summary, genomic testing is assuming an increasingly prominent role in the identification, prioritization, and interpretation of disease-associated genomic variants. While a few of these variants are known to be pathogenic, knowledge of the deleterious effects of the majority remain elusive. Laboratory methods remain gold-standards for functional characterization, but are generally incompatible with large-scale characterization of variant effects. Predictive algorithms are in some instances successfully applied to differentiate pathogenic variants from variants of unknown significance. However, these methods largely ignore measures of protein structure, energies, molecular bonds, intermolecular interactions, post-translational modification effects, protein aggregation, and stability. Conversely, this information lies at the heart of molecular modeling and dynamic simulation, which collectively equip us more fully to grapple with interpreting the effects of VUSs. We strongly believe that these methodologies constitute the future frontiers in forging the analytical pipeline of interpreting the results of genomic testing for individualized medicine and advocate their increased deployment within variant characterization efforts.",
"appendix": "Author contributions\n\n\n\nEWK & GRO conceived the manuscript. GRO wrote the manuscript. MTZ generated figures and wrote the manuscript. RAU wrote the manuscript. All authors proof-read and approved the manuscript and have been involved in the modeling and interpretation of genomic variants in diagnostic odyssey, hematology and oncology studies within Mayo Clinic’s Center for Individualized Medicine.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe acknowledge Mayo Clinic Center for Individualized Medicine for making the work described possible.\n\n\nReferences\n\nCollins FS, Varmus H: A new initiative on precision medicine. N Engl J Med. 2015; 372(9): 793–795. PubMed Abstract | Publisher Full Text\n\nYang Y, Muzny DM, Xia F, et al.: Molecular findings among patients referred for clinical whole-exome sequencing. JAMA. 2014; 312(18): 1870–1879. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang YP, Muzny DM, Reid JG, et al.: Clinical whole-exome sequencing for the diagnosis of mendelian disorders. N Engl J Med. 2013; 369(16): 1502–1511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichards S, Aziz N, Bale S, et al.: Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015; 17(5): 405–424. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartí-Renom MA, Stuart AC, Fiser A, et al.: Comparative protein structure modeling of genes and genomes. Annu Rev Biophys Biomol Struct. 2000; 29: 291–325. PubMed Abstract | Publisher Full Text\n\nSali A: Modeling mutations and homologous proteins. Curr Opin Biotechnol. 1995; 6(4): 437–451. PubMed Abstract | Publisher Full Text\n\nZhang Y: Protein structure prediction: when is it useful? Curr Opin Struct Biol. 2009; 19(2): 145–155. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIhle NT, Byers LA, Kim ES, et al.: Effect of KRAS oncogene substitutions on protein behavior: implications for signaling and clinical outcome. J Natl Cancer Inst. 2012; 104(3): 228–239. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuffy MR, Parker AL, Kalkman ER, et al.: Identification of novel small molecule inhibitors of adenovirus gene transfer using a high throughput screening approach. J Control Release. 2013; 170(1): 132–40. PubMed Abstract | Publisher Full Text\n\nSchmidt T, Bergner A, Schwede T: Modelling three-dimensional protein structures for applications in drug design. Drug Discov Today. 2014; 19(7): 890–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHillisch A, Pineda LF, Hilgenfeld R: Utility of homology models in the drug discovery process. Drug Discov Today. 2004; 9(15): 659–669. PubMed Abstract | Publisher Full Text\n\nKelley LA, Mezulis S, Yates CM, et al.: The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc. 2015; 10(6): 845–858. PubMed Abstract | Publisher Full Text\n\nBordoli L, Kiefer F, Arnold K, et al.: Protein structure homology modeling using SWISS-MODEL workspace. Nat Protoc. 2009; 4(1): 1–13. PubMed Abstract | Publisher Full Text\n\nCosta MG, Benetti-Barbosa TG, Desdouits N, et al.: Impact of M36I polymorphism on the interaction of HIV-1 protease with its substrates: insights from molecular dynamics. BMC Genomics. 2014; 15(Suppl 7): S5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKovalskyy D, Dubyna V, Mark AE, et al.: A molecular dynamics study of the structural stability of HIV-1 protease under physiological conditions: the role of Na+ ions in stabilizing the active site. Proteins. 2005; 58(2): 450–8. PubMed Abstract | Publisher Full Text\n\nTorbeev VY, Raghuraman H, Hamelberg D, et al.: Protein conformational dynamics in the mechanism of HIV-1 protease catalysis. Proc Natl Acad Sci U S A. 2011; 108(52): 20982–7. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "13600",
"date": "06 May 2016",
"name": "Saverio Alberti",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article the Authors seek ways on how to better interpret the functional impact of mutations/polymorphisms currently classified as variants of unknown significance (VUSs). To tackle such a problem, the authors propose “molecular modeling and molecular dynamic simulations as tools that can significantly increase the yield of information to aid in the evaluation of pathogenicity.” Expectations are that this “will significantly benefit the mission of clinical genomics for individualized medicine.”The idea is good and deserves to be published. Most of all, it deserves to be extensively discussed. Points to be discussedAs much as the idea is good, it would benefit from better highlighting its (current) limitations. Ab initio molecular modeling is a difficult enterprise. It can rarely be performed to some extent of accuracy by automatized engines/softwares and often reaches good structure prediction in around 60% of cases.Several large-scale, tentatively whole-coding genome, protein crystallization initiatives have been undertaken. Discussion of the state of the art of these initiatives may provide metrics on what could be expected from real-life application of the Authors’ proposal, and how this is expected to evolve in the near future. Most mutations/polymorphisms are detected in non-coding regions of DNA. This is hardly surprising given that non-coding DNA represents approximately 97% of the total content of a mammalian cell. The authors mention mutations in non coding regions, but this deserves being more extensively discussed in the publication. SuggestionsDiscussion of the points above and of the suggestions below, may possibly lead to extend the proposed approach to additional layers of prediction.Algorithms are there that do recognise promoter, enhancer and super-enhancer regions with increasing accuracy. Clearly, allocating a mutation/polymorphism in one such region would increase its investigational and possibly medical value.Additional regions of interest are CpG islands upstream or, less reliably, within coding regions.Additional regions of interest are anchoring sites of DNA loops to distinct nuclear/nuclear membrane regions.Additional regions of interest are those hosting the increasing family of non-coding RNA (NCRNA), including long NCRNA and miRNA, that have been shown to play distinct regulatory functions and may have roles in cancer development.",
"responses": []
},
{
"id": "13880",
"date": "18 May 2016",
"name": "Attila Bérces",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors of this article \"highlight a role for molecular modeling and molecular dynamic simulations as tools that can significantly increase the yield of information to aid in the evaluation of pathogenicity.\"\nMolecular dynamics simulations are indeed excellent tools to study protein and nucleic acid structures, protein-ligand, protein-protein and protein-nucleic acid interactions. These structural changes and interactions are fundamental in molecular biology and thus molecular simulations are an excellent tool to increase our understanding of these biological processes. In particular the proposed method may help the understanding of coding variants associated with amino acid substitution. In addition, RNA structure can be modeled, which can be informative about some micro-RNA variants.\nHowever, these methods are not without limitations and this paper does not state these limitations explicitly. Pathogenic effects come in all strength and complexities. Most often pathogenicity is not associated with a single variant, but with a combination of several variants and the effect of a particular variant depends on the genotypes of other genes. Most likely, variants will stay of unknown significance unless we develop a system level understanding of biology. We may reach a point that we can simulate viral functions at molecular level, but it will not be applied to human variants any time soon.\nMolecular modelling has some limitations and requires expert knowledge to apply it properly.\nWhile I agree that molecular modelling is an excellent tool to understand molecular biology, we need to develop a system level of understanding of the connection between molecular biology and genetics before we can apply molecular simulations. Currently, molecular modeling can make practical contribution to understand variants associated with strong effects or the molecular basis of some Mendelian disorders.",
"responses": []
},
{
"id": "13603",
"date": "31 May 2016",
"name": "Anita Grigoriadis",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article Oliver et al. highlight to importance of analytical methods such as “molecular modeling and molecular dynamic simulations\" as tools that can significantly increase the interpretation of the functional impact of mutations/polymorphisms currently classified as variants of unknown significance (VUSs).\nWhile the topic of molecular dynamics simulation for VUSs is worth discussing, the article could expand on\nthe current limitation of these methods the complexity of these nucleotide changes in e.g. coding and non-coding genomic regions integration with other high throughput technologies and thus increase challenges for multimodal effects in such simulations HIV is given as an example, however other clinical examples, their associated specificity and limitations should be provided to make this article relevant for a wider audience.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-766
|
https://f1000research.com/articles/5-1778/v1
|
21 Jul 16
|
{
"type": "Research Article",
"title": "Does sadness impair color perception? Flawed evidence and faulty methods",
"authors": [
"Alex O. Holcombe",
"Nicholas J. L. Brown",
"Patrick T. Goodbourn",
"Alexander Etz",
"Sebastian Geukes",
"Alex O. Holcombe",
"Patrick T. Goodbourn",
"Alexander Etz",
"Sebastian Geukes"
],
"abstract": "In their 2015 paper, Thorstenson, Pazda, and Elliot offered evidence from two experiments that perception of colors on the blue–yellow axis was impaired if the participants had watched a sad movie clip, compared to participants who watched clips designed to induce a happy or neutral mood. Subsequently, these authors retracted their article, citing a mistake in their statistical analyses and a problem with the data in one of their experiments. Here, we discuss a number of other methodological problems with Thorstenson et al.’s experimental design, and also demonstrate that the problems with the data go beyond what these authors reported. We conclude that repeating one of the two experiments, with the minor revisions proposed by Thorstenson et al., will not be sufficient to address the problems with this work.",
"keywords": [
"Mood",
"perception",
"color",
"open data",
"reanalysis"
],
"content": "Introduction\n\nBased on two experiments, Thorstenson et al. (2015a) claimed that a state of sadness—induced by watching a short film clip—impairs performance on a specific perceptual task: discrimination of colors along the blue–yellow axis, but not the red–green axis. This conclusion is interesting because it is specific to a single dimension of color space; poor performance on tasks generally, or low willingness to cooperate with an experimenter, would not be a surprising effect of sadness.\n\nIn their retraction notice (Thorstenson et al., 2015b), the authors acknowledged that their data did not justify their conclusion that impairment was specific to one aspect of color space. They also described an anomaly in the histogram of the data of Experiment 2. In our sections below entitled “A confounded comparison” and “Perceptual impairment or change in bias?”, we detail other problems with the way the experiments were conducted, the choice of stimuli, and the measures chosen. It is these problems that lead us to believe that even the revised Experiment 2 (proposed by Thorstenson et al., 2015b) will not justify their original conclusion. In our “Re-analysis of data” section, we describe a number of statistical issues that go beyond the problem mentioned in the retraction notice. In our final section, “Anomalies and strange patterns in the data”, we report some other anomalies with the dataset, which further undermine confidence in the way the experiments were carried out and in the resulting findings. We offer this analysis not only to improve the state of the literature on the perceptual effects of watching sad film clips, but also in the hope that it will lead to better work in this area in the future.\n\n\nA confounded comparison\n\nWhen designing an experiment comparing two conditions, one strives to make the factor of interest the only difference between the conditions. Thorstenson et al. (2015a) contrasted two film clips, one of which was intended to cause the participants to feel sad. The clips ought to have been chosen to avoid any other differences (on average) in their effect on the participants. The “sadness” clip of Experiment 1 is an excerpt from the animated Disney movie The Lion King, with an unusual lighting that gives the impression of daylight filtered through dust, while the “happiness” clip is a warmly-lit, indoor recording of the comedian Bill Cosby. The “sadness” clip used in Experiment 2 is the same Lion King excerpt, converted from color to grayscale, and the “neutral” clip is a grayscale film of sticks appearing on top of one another at different orientations (also converted from color to grayscale).\n\nUnfortunately, differences in mean color and the color variability in these clips may have differently affected subsequent perception of blue and yellow versus red and green. For example, the contrast along the blue–yellow axis might have been greater in the sadness clips. Such a difference would result in reduced sensitivity to blue–yellow (Krauskopf et al., 1982). The use of grayscale clips does not eliminate this issue. An analysis of the Hue, Saturation, Brightness (HSB) values in the movie files posted online indicates that the mean color of the grayscale clips is bluish-reddish, with some saturations approaching 5%. These grayscale clips, therefore, may have had differences in average color as well as color contrast that were uncontrolled. To resolve the issue, the colors displayed on the laboratory screen must be measured with a colorimeter. The authors should have made these measurements and reported them in their paper, in order to provide an indication of whether simple contrast adaptation specific to each color axis would occur when viewing the clips. In the absence of a report of these measurements or any mention of the issue, it appears that Thorstenson et al. (2015a) did not take the appropriate steps to eliminate the possibility that a classic process in color perception could explain the results.\n\nBlue, yellow, green and red are all defined relative to a white point that, in the human visual system, is quite flexible. Just as one can adjust the white balance of a camera to fit scenes with different illumination, for humans the point considered to be the center of color space changes depending on the palette of colors that confronts us (Webster & Leonard, 2008). Unfortunately, Thorstenson et al. displayed their test stimuli in a manner unsuited to controlling the participant’s white point. Color perception experiments typically use a neutral grey or white background to provide a white-point reference for participants, alongside the test stimulus. Thorstensen et al.’s use of full-field color without a simultaneous reference stimulus makes categorization of desaturated patches problematic. In such circumstances, the participants’ white points may be more dependent on the color content of the movie clip they viewed previously, which as mentioned above appears to have been uncontrolled. In addition, the lack of a grey or white reference stimulus may cause participants to be completely unable to judge the stimulus color more often. In such circumstances, responses may be particularly prone to influence by cognitive factors or by priming (García-Pérez & Alcalá-Quintana, 2013).\n\nIn addition to color, there may be other confounding difference between the two types of clips. The clips likely differed in interest, action, and other features. Unfortunately, it is difficult to know whether such features might have affected participants’ color perception. It certainly is possible for such differences to bias the participants’ responses when they are uncertain of the stimulus’s color. Of course, it is almost impossible to avoid featural differences between any two particular clips. Because of this, a good experimental design would utilize a large set of clips, assess the various featural differences between the stimuli, and either match the two groups of clips carefully on their features, or model them as random effects in a mixed-effects model (Wells & Windschitl, 1993).\n\n\nPerceptual impairment or change in bias?\n\nThorstenson et al. (2015a) concluded that sadness “impair[s] color perception on the blue–yellow color axis” (p. 1). But signal detection theory, which was not used, would be necessary to show whether the decrease in accuracy found was indeed due to an impairment in color perception (i.e., a decline in sensitivity along the blue–yellow axis), or whether the judgments of the sadness group were instead biased away from blue and yellow. For decades, studies of perception have used signal detection theory to distinguish between a change in perceptual ability and a change in, say, cognitive bias to press the blue or yellow button rather than the red or green one (Green & Swets, 1966). Unfortunately, Thorstenson et al.’s plan to simply repeat Experiment 2 with minor revisions would not allow for the appropriate analysis. In Experiment 2, participants were tested in only two trials for each stimulus. Much more data would be needed to discern between a decline in the participants’ ability to discriminate the colors from a decline in the participants’ bias toward pressing the blue or yellow button (instead of the red or green). For the four-alternative categorization task used by Thorstenson et al., a multivariate extension of signal-detection theory should be used (such as general-recognition theory, Ashby & Townsend, 1986).\n\nThe analyses of Thorstenson et al. (2015a), and also the improved analyses that we have suggested above, assume stochastic independence of participants’ accuracies on the red–green stimuli and the blue–yellow stimuli. Unfortunately, however, this assumption may be unjustified. Participants’ accuracy on one axis might affect their guessing strategy on another. In Experiment 1 for example, accuracy was very high on the blue–yellow axis, suggesting that many participants may have had a clear color percept of the blue or yellow stimuli, but were less certain about the red and green stimuli. If so, when an unclear patch came up and they guessed, they may have been unlikely to guess blue or yellow, in an effort to balance their responses across the available options (many participants may have correctly guessed that the stimuli were roughly equally distributed among the four categories). This would artifactually improve performance on the red and green stimuli. Modeling this phenomenon, however, would be difficult. Even if we had access to the raw responses (rather than the summary data provided by Thorstenson et al.), it would be difficult to estimate the participants’ guessing strategy. To avoid this problem in a future version of this experiment, we suggest that Thorstenson and colleagues should consider adopting the two-alternative forced choice design (Fechner, 1860/1966) commonly used in psychophysics.\n\nThorstenson et al. (2015a) are not the only researchers to have used bias-prone measures of perception to support claims that some non-perceptual state can influence perception. Firestone & Scholl (2015) provide many other examples, with useful discussion.\n\n\nReanalysis of data\n\nThere are issues with the dataset (Thorstenson et al., 2015c) that were not described in the retraction (Thorstenson et al., 2015b) of the article. Some of these issues affect Experiment 1, which Thorstenson et al. indicated that they plan to re-publish. We describe and discuss these issues in this section, as well as the next section, “Anomalies and strange patterns in the data”.\n\nThe most important empirical claim made by Thorstenson et al. (2015a) was that there was a difference in performance between their two measures, namely color perception along the blue–yellow axis and color perception along the red–green axis. However, these authors provided no statistical test of a difference in the effect of the film clip on blue–yellow compared to red–green. This problem was discussed widely on blogs and on PubPeer (2016), and was acknowledged by Thorstenson et al. (2015b) in their retraction notice. Thorstenson et al. (2015a, p. 4) noted that the possible difference between red–green and blue–yellow color perception, such that “sadness influenced chromatic judgments about colors on the blue–yellow axis, but not those on the red–green axis,” is critical to ruling out “the possibility that sadness simply led to less effort, arousal, attention, or task engagement.” Such a difference implies a statistical interaction between the “emotion condition” and “color axis” factors. However, the authors did not report a statistical test for this interaction in either of their experiments. When we (and the authors of various blogs, such as Areshenkoff, 2015) tested this interaction with the published data, we found (code at: Holcombe et al., 2016) that it was not statistically significant: Experiment 1, F(1, 125) = 3.51, p = .06; Experiment 2: F(1, 128) = 0.40, p = .52. In their retraction notice, Thorstenson et al. (2015b) reported a z test to test the same issue (for unknown reasons, they did not use a conventional statistical interaction, but instead followed a procedure described in Rosenthal & Rosnow, 1991), which also did not yield statistical significance.\n\nAn additional potential source of error is that Thorstenson et al. (2015a) did not record the color-perception performance of their participants before the film clips were shown. It was apparently considered sufficient to randomize the participants to watch one of two film clips; presumably the reasoning was that this randomization made it unlikely that the two groups differed much in baseline performance. However, even if this assumption were to be confirmed, the two groups would likely differ somewhat at baseline, even if by only a small amount, and such a difference could have an effect on the outcome given the relatively small sample sizes involved (Saint-Mont, 2015). It would have been useful for these differences to be measured and included in the subsequent analyses, given that Thorstenson et al.’s hypothesis was that sadness would “impair” (i.e., reduce, compared to a previous state) participants’ color perception. In addition, using a change score for each participant can increase statistical power by reducing the contribution of variation among participants to the error term.\n\nFinally, we note that Thorstenson et al.’s (2015a) experimental design assumes the complete independence of participants’ accuracy on the two sets of stimuli (red–green and blue–yellow). We discuss a possible violation of this assumption in our “Perceptual impairment or change in bias?” section above.\n\n\nAnomalies and strange patterns in the data\n\nWe observed a strange pattern in the data for the blue–yellow axis in Thorstenson et al.’s (2015a) Experiment 2. Specifically, a very large number of participants (53 out of 130) had a score of exactly 50%, corresponding to 12 out of 24 correct responses, with every other number of correct responses (10, 11, 13, 14, etc.) being achieved by a much smaller number of participants. This is illustrated in Figure 1, where the spike at the 50% level is clearly visible. This issue was one of the reasons given by Thorstenson et al. (2015b) for retracting their article.\n\nThe histogram shows the number of occurrences of each score, calculated by Thorstenson et al. (2015a) as the proportion of correct responses in the 24 trials in which either a blue or yellow patch was presented. The pairs of adjacent bars near 0.55 and 0.8 correspond to cases that are not compatible with correct rounding.\n\nUpon our request, Christopher Thorstenson provided us with per-color patch data. The per-color patch data consists of two Excel files (one per experiment), with each cell containing a combined score for the participants’ two responses for each color and saturation level. The score for each case is either 0.0, 0.5, or 1.0, corresponding to 0, 1, or 2 correct responses (see https://osf.io/sbhn9/).\n\nCloser examination of the per-color patch data, shows that of the 53 participants scoring exactly 50%, 49 (i.e., 37.7% of all participants in Experiment 2) had identical scores for both colors, namely 6.0 (100%) for blue and 0.0 (0%) for yellow (in their patch data files, each correct observation counts for a half-point, so that scores for each color range from 0.0 to 6.0 in increments of 0.5; thus, a score of 6.0 corresponds to 12 correct responses out of 12). We are at a loss to explain this phenomenon, which affected both experimental conditions (26 of the 49 participants with this 12–0 split were in the neutral condition, with 23 of the 49 being in the sadness condition). There seems no reason to suppose that the undergraduate participants in this experiment would have been markedly less sensitive to yellow than those in Experiment 1. However, even if their ability to distinguish the color yellow was affected by some environmental factor, or if they had been accidentally (perhaps due to a software problem) shown, say, a gray patch instead of a yellow one, their expected score for yellow would be 1.5 (i.e., three correct identifications out of 12 attempts) by chance alone.\n\nA further concern with the summary data (Thorstenson et al., 2015c) is that the conversion of color perception values from counts of responses to percentages of correct attempts for both axes in Experiment 2 appears to be inconsistent. These percentage values, reported to two decimal places, ought to be the result of dividing the number of successful attempts on each axis (i.e., the total number of correct identifications of red or green patches for the red–green axis, and the total number of correct identifications of blue or yellow patches for the blue–yellow axis) by 24. For example, an examination of the patch scores shows that participants #4 and #5 both scored a total of 6.5 for blue and yellow patches combined, corresponding to 13 correct identifications out of 24 on the blue–yellow axis. However, in the published dataset file, participant #4 has a value of 0.54 for the corresponding percentage variable BY_ACC (blue–yellow accuracy), whereas participant #5 has a value of 0.55 for the same variable (the true value of 13/24 being 0.54166¯). Christopher Thorstenson (personal communication, December 1, 2015) has explained to us that these percentages resulted from taking the mean of the individual percentages of correct attempts for each color of the axis in question (e.g., red and green), with these individual percentages having first been rounded to two decimal places. It is not clear whether this explains all the anomalies in the data for Experiment 2; in any case, it serves as a reminder that, in order to avoid loss of information, rounding should be avoided during an analysis and only applied, if necessary, during the final reporting of results.\n\nAn examination of the distribution of the scores for the two color axes reveals considerable differences between Experiment 1 and Experiment 2. In Experiment 1, the distribution for both axes was substantially negatively skewed, with the majority of participants correctly identifying almost all of the patches for all four colors (Figure 2a). In Experiment 2, the score distribution was different for each axis. For the red–green axis (Figure 2b, top panel) the scores were approximately normally distributed: roughly similar numbers of participants achieved each possible score, with a small number having very low or very high scores. In contrast, the blue–yellow axis was positively skewed, displaying the “spike” discussed previously (Figure 2b, bottom panel).\n\nDistribution of blue–yellow and red–green axis scores in (a) Experiment 1 and (b) Experiment 2. Histograms show the number of occurrences of each score for the red–green axis (top panels) and blue–yellow axis (bottom panels). The range of scores on the x-axis is 0–12, reflecting Thorstenson et al.’s (2015a) scoring scheme of 0.5 points per correct answer, with 12 trials per color and two colors per axis. Note the discontinuity in the y-axis for blue–yellow in Experiment 2 (bottom-right panel), added to accommodate the surprisingly high peak at 6.\n\nUsing the patch-level data, we broke the two-color axis scores down into individual colors, as shown in Figure 3. For Experiment 1, the per-color data more or less followed the pattern of the two-color axis of which each color was a part (Figure 3a); this was also true for the red–green axis in Experiment 2 (Figure 3b, left panels). However, an even stranger pattern emerged for the blue–yellow axis in Experiment 2 (Figure 3b, right panels). Of the 130 participants, 106 (81.5%) scored a maximum 6.0 (corresponding to 12 correct responses) for blue, while 56 (43.1%) scored zero for yellow. The observed “spike” at 50% (i.e., 12 out of a possible 24 correct responses) for the blue–yellow axis is thus mostly explained by people who had a perfect score (12 out of 12) for blue, while completely failing to recognize yellow patches at any saturation and thus obtaining a score of 0.\n\nDistribution of color patch scores in (a) Experiment 1 and (b) Experiment 2. Histograms show the number of occurrences of each score for the red, green, blue and yellow color patches. The range of scores on the x-axis is 0–6, reflecting Thorstenson et al.’s (2015a) scoring scheme of 0.5 points per correct answer, with 12 trials per color. Note the discontinuity in the y-axis for blue patches in Experiment 2 (bottom row, second panel from right), added to accommodate the surprisingly high peak at 6.\n\nFigure 4 plots the participants’ performance for each color, broken down further into the proportion of correct responses for each saturation level (recall that participants were asked to identify colors at each of six different levels of saturation.) In Experiment 1, this resulted in what appears to be a ceiling effect – mean accuracy reaches 90% or more already at the third-lowest color saturation level (.10) and levels off as saturation increased thereafter (Figure 4a). In Experiment 2, the ceiling effect disappeared for the red–green axis, for which scores on both colors improved approximately linearly with increasing color saturation (Figure 4b, left panels); however, on the blue–yellow axis, the effect of the split between the two colors is once again clear. The ceiling effect is even more pronounced for blue here than in Experiment 1, while scores for yellow are low even at the highest color saturation level (Figure 4b, right panels).\n\nColor categorization as a function of saturation in (a) Experiment 1 and (b) Experiment 2. Each panel shows the proportion of correct responses for each of the six saturation levels for a given color. Mean performance in the sadness condition is represented by triangles joined by solid lines, and mean performance in the happiness (Experiment 1) and neutral (Experiment 2) conditions is represented by circles joined by dashed lines. Error bars show parametric 95% confidence intervals on the means. (Details of the color calibration procedure were not stated by Thorstenson et al. (2015a), so it is not clear how to interpret these saturation values.)\n\nIt is difficult to imagine what might have caused these results in Experiment 2. The Method section for this experiment suggests that the only change that was made from Experiment 1 was the nature of the film clips that were shown to participants. The differences for both axes (and, indeed, for all four colors) between Experiments 1 and 2—regardless of the film clip watched by participants—are puzzling, given that both samples were drawn from the same population of undergraduates and hence ought not to differ widely in their physiological characteristics. Because the color characteristics of the two sets of film clips were apparently not well-controlled, one possible explanation for this discrepancy is differential adaptation of the color mechanisms in the visual system, which adds to our concern about possible confounds (see our section “A confounded comparison?”). But we have difficulty believing that this, on its own, could account for such a substantial difference between the two experiments.\n\nGiven that the extreme blue–yellow scores in Experiment 2 were obtained from participants in both the neutral and sadness conditions, a further possibility is that simply watching grayscale film clips for a few minutes was sufficient to substantially distort participants’ color vision (on the blue–yellow axis only). However, if Thorstenson and colleagues had noticed such a finding, they would presumably have mentioned it in their article (and perhaps alerted colleagues in the field of physiology to this remarkable discovery). Otherwise, we are left with two possible conclusions: either around 40% of the participants in Thorstenson et al.’s (2015a) Experiment 2 all had the same problem with their vision (which was not shared by any of the participants in Experiment 1), or some form of equipment failure or other technical problem caused this unusual pattern of values to be recorded. In any case, it seems likely that Thorstenson et al. failed to notice this anomaly when examining their data prior to performing their statistical analyses.\n\n\nConclusion\n\nWhile we strongly support the retraction by Thorstenson et al. (2015b) of their article (Thorstenson et al., 2015a) on the basis of the problems they noted with Experiment 2, we maintain that the basic methodology of both of their experiments is flawed. As Thorstenson and colleagues move forward, together with others who seek to assess whether mood and other factors can influence perception, they should bring their work up to modern standards of statistics and psychophysics. Doing so for experiments like those of Thorstenson et al. would involve: (1) careful control of the visual differences between the movie clips, or, better, mood induction via non-visual stimuli such as an audio recording of a story; (2) the use of many movie clips or recordings, and mixed-effects analysis to address differences that cannot be eliminated between any two clips or recordings; (3) a baseline measurement of color perception; (4) an analysis based on signal-detection theory.\n\n\nData availability\n\nOpen Science Framework: Reanalysis of Thorstenson et al.’s (2015) “Sadness Impairs Color Perception”, doi 10.17605/osf.io/kwuq4 (Brown et al., 2016).\n\nWe have archived the R code that we used to analyze the data and generate our figures at the Open Science Framework (OSF; doi: 10.17605/osf.io/kwuq4). This code works with the original dataset files uploaded to OSF (Thorstenson et al., 2015c), together with the patch data files that Christopher Thorstenson sent us (by “patch data”, we mean data broken down to the individual color patches tested) and that we posted at https://osf.io/sbhn9/ (Thorstenson subsequently asked us to delete two of the files, which we did).",
"appendix": "Author contributions\n\n\n\nAOH coordinated the team and wrote most of the main section of the article. NJLB started the discussion (on Twitter), wrote the R code to perform the detailed analysis of the dataset, and wrote most of the sections describing this analysis. PTG contributed several points and Figure 4 as well as contributing to the writing. AE contributed to the discussion of stimuli problems and (lack of) necessary stimulus sampling. SG contributed to the analysis, contributed early versions of some of the R code, and helped revise the manuscript. All authors agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nAOH is an associate editor for a journal, Perspectives on Psychological Science, that is published by the organization (the Association for Psychological Science) that also publishes the journal in which Thorstenson et al. (2015) appears.\n\n\nGrant information\n\nThis work was supported by internal university grants to AOH and PTG from the University of Sydney and the University of Melbourne, respectively. No other authors received any private or public funding to support their involvement in this work.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Christopher Thorstenson for sharing the per-color patch data, and for his comments on a previous version of this manuscript.\n\n\nReferences\n\nAreshenkoff CN: On the importance of plotting; or — Psych. Science will publish anything. [weblog post]. 2015. Reference Source\n\nAshby FG, Townsend JT: Varieties of perceptual independence. Psychol Rev. 1986; 93(2): 154–179. PubMed Abstract | Publisher Full Text\n\nBrown NJL, Holcombe AO, Etz A, et al.: Reanalysis of Thorstenson et al.’s (2015) “Sadness Impairs Color Perception”. Open Science Framework. 2016. Data Source\n\nFechner GT: Elements of psychophysics. New York, NY: Holt, Rinehart and Winston. (Original work published 1860). 1966. Reference Source\n\nFirestone C, Scholl BJ: Cognition does not affect perception: Evaluating the evidence for ‘top-down’ effects. Behav Brain Sci. 2015; 20: 1–77. PubMed Abstract | Publisher Full Text\n\nGarcía-Pérez MA, Alcalá-Quintana R: Shifts of the psychometric function: distinguishing bias from perceptual effects. Q J Exp Psychol (Hove). 2013; 66(2): 319–37. PubMed Abstract | Publisher Full Text\n\nGreen DM, Swets JA: Signal Detection Theory and Psychophysics. New York, NY: Wiley. 1966. Reference Source\n\nHolcombe AO, Brown NJL, Etz A, et al.: Code. 2016. Publisher Full Text\n\nKrauskopf J, Williams DR, Heeley DW: Cardinal directions of color space. Vision Res. 1982; 22(9): 1123–1131. PubMed Abstract | Publisher Full Text\n\nPubPeer: Sadness impairs color perception. 2016. Reference Source\n\nRosenthal R, Rosnow RL: Essentials of Behavioral Research: Methods and Data Analysis. New York, NY: McGraw Hill. 1991. Reference Source\n\nSaint-Mont U: Randomization Does Not Help Much, Comparability Does. PLoS One. 2015; 10(7): e0132102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThorstenson CA, Pazda AD, Elliot AJ: Sadness impairs color perception. Psychol Sci. 2015a. Publisher Full Text\n\nThorstenson CA, Pazda AD, Elliot AJ: Retraction of “Sadness impairs color perception”. Psychol Sci. 2015b; 26(11): 1822. [The publisher has given the retraction the same DOI as the original article, perhaps accidentally.]. Publisher Full Text\n\nThorstenson CA, Pazda AD, Elliot AJ: Sadness impairs color perception. [Data and stimuli set]. Republished by Holcombe et al. (2016). 2015c. Publisher Full Text\n\nWebster MA, Leonard D: Adaptation and perceptual norms in color vision. J Opt Soc Am A Opt Image Sci Vis. 2008; 25(11): 2817–2825. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWells GL, Windschitl PD: What's in a question? Contemp Psychol. 1993; 38(4): 383–385. Publisher Full Text"
}
|
[
{
"id": "15370",
"date": "02 Aug 2016",
"name": "Sophie von Stumm",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article reviews a previous publication that was based on two flawed experiments and has since been retracted. The authors thoroughly analyse the experimental methods that were employed in the retracted study, and they support the retraction. In addition, they highlight further problems with the experimental design and they make specific recommendations for improving future studies in this area.",
"responses": []
},
{
"id": "15140",
"date": "08 Aug 2016",
"name": "Michael A Crognale",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript by Holcombe et al. as a comment on Thorstenson et al.'s prior manuscript is well well written and reasoned. The problems with the Thorstenson et al. article are many and most of them have been well addressed here. However, since the manuscript by Throstenson et al. has been retracted reportedly for other reasons, it does not seem particularly fruitful to critique the original manuscript further unless a resubmission has published. On the other hand, pointing out the additional methodological and statistical problems with the original manuscript may be instructive for those following up on this work as suggested by Holcomb et al. In particular, there does seem to be a disturbing trend to attribute results to factors that have not been well established as uniquely or even likely causal, (e.g. in the present case that \"sadness\" caused the observed trends in the data). Holcomb et al.'s manuscript also illustrates the importance of scrutinizing the raw data for evidence of faulty methodology, and ceiling/floor effects. Perhaps publishing the present manuscript may be worthwhile as it is a potentially valuable instructional tool.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1778
|
https://f1000research.com/articles/5-1763/v1
|
20 Jul 16
|
{
"type": "Research Note",
"title": "How much is a pheromone worth?",
"authors": [
"Jose Mauricio S. Bento",
"Jose Roberto P. Parra",
"Silvia H. G. de Miranda",
"Andrea C. O. Adami",
"Evaldo F. Vilela",
"Walter S. Leal",
"Jose Mauricio S. Bento",
"Jose Roberto P. Parra",
"Silvia H. G. de Miranda",
"Andrea C. O. Adami",
"Evaldo F. Vilela"
],
"abstract": "Pheromone-baited traps have been widely used in integrated pest management programs, but their economic value for growers has never been reported. We analyzed the economic benefits of long-term use of traps baited with the citrus fruit borer Gymnandrosoma aurantianum sex pheromone in Central-Southern Brazil. Our analysis show that from 2001 to 2013 citrus growers avoided accumulated pest losses of 132.7 million to 1.32 billion USD in gross revenues, considering potential crop losses in the range of 5 to 50%. The area analyzed, 56,600 to 79,100 hectares of citrus (20.4 to 29.4 million trees), corresponds to 9.7 to 13.5% of the total area planted with citrus in the state of São Paulo. The data show a benefit-to-cost ratio of US$ 2,655 to US$ 26,548 per dollar spent on research with estimated yield loss prevented in the range of 5-50%, respectively. This study demonstrates that, in addition to the priceless benefits for the environment, sex pheromones are invaluable tools for growers as their use for monitoring populations allows rational and reduced use of insecticides, a win-win situation.",
"keywords": [
"Gymnandrosoma aurantianum",
"citrus fruit borer",
"benefit-cost analysis",
"pest management",
"monitoring"
],
"content": "Introduction\n\nThe discovery of bombykol as the sex pheromone of a domesticated insect species (Butenandt et al., 1959) triggered the interest of entomologists and natural product chemists to jointly identify pheromones from economically important insect pests and explore their potential in Integrated Pest Management (IPM) (Howse et al., 1998; Ridgway et al., 1990; Silverstein, 1981). This interest continues to increase to date given the need for environmentally friendly alternatives to control insect pest populations. After all, pheromones are non-polluting and usually non-toxic natural products. Strictly speaking they are nature-inspired synthetic compounds, i.e., identical to natural products, but manmade chemical signals. Additionally, pheromones are species-specific and safe for beneficial organisms; thus, they are ideal components of IPM programs (Jutsum & Gordan, 1989). Of note, pheromones have been registered in many countries for use in pest management, and no evidence of adverse effects has been reported (Witzgall et al., 2010). There is a consensus that successful implementation of pheromones in the field frequently involves a joint effort by chemical ecologists, entomologists and/or extension agronomists, and growers, in addition to the pheromone industry (Witzgall et al., 2010).\n\nThere are many ways in which pheromones can be used for surveillance and IPM programs, including monitoring, attract-and-kill, and mating disruption. Pheromone-baited traps are sufficiently sensitive to detect low population densities and are therefore an effective way for tracking invasive species while they are still at the establishment stage (El-Sayed et al., 2006; Liebhold & Tobin, 2008). Population monitoring has been a simple and widely used strategy to determine the ideal moment for application of control procedures (i.e., insecticides), using pre-defined thresholds (action levels) based on level of capture in pheromone-baited traps. This strategy reduces the use of insecticides to the minimal amount necessary to protect both crops and the environment (Thomson et al., 1999; Witzgall et al., 2010).\n\nOne of the first systems to use an action level based on capture with pheromone-baited traps was established for the pea moth Cydia nigricana (Wall et al., 1987). Later, many other studies were conducted with equal success in agricultural, horticultural or forestry applications, against pest species including the European corn borer Ostrinia nubilalis (Laurent & Frérot, 2007), tufted apple budmoth Platynota idaeusalis (Knight & Hull, 1989), lightbrown apple moth (Bradley et al., 1998), scale insects (Dunkleblum, 1999), Mullein bug Campylomma verbasci (McBrien et al., 1994), grapevine moth Lobesia botrana (Ioriatti et al., 2011), codling moth Cydia pomonella (Madsen & Vakenti, 1973), Oriental fruit moth Grapholita molesta (Rothschild & Vickers, 1991), pink bollworm Pectinophora gossypiella (Qureshi et al., 1984), Old World bollworm Helicoverpa armigera (Cameron et al., 2001), cotton leafworm Spodoptera litura (Singh & Sachan, 1993), and yellow rice stem borer Scirpophaga incertulas (Krishnaiah et al., 1998), just to cite a few.\n\nDespite the clear advantages offered by the use of pheromones in IPM in recent decades, particularly the use of traps for monitoring in extensive areas, to date there are no studies on their economic benefits. While the benefits for the environment are less tangible, the economic benefits could be estimated. Evidence of economic benefits could be extremely helpful in motivating growers to employ environmentally friendly strategies for pest control, the chemical industry to participate in production and commercialization of pheromones, and the public and private sector to promote and support more translational research.\n\nThe citrus fruit borer Gymnandrosoma aurantianum Lima (Lepidoptera, Tortricidae) is a representative case for analysis of the economic benefits achieved by the use of a synthetic pheromone to manage this pest in extensive areas. Brazil is the leading worldwide producer of citrus (USDA, 2015), and Central-Southern Brazil, an area with generalized occurrence of the citrus fruit borer, accounts for approximately 80% of all citrus production in the country (IBGE, 2015). Females normally deposit a single egg per fruit (Garcia & Parra, 1999); after eclosion, the larvae pierce the skin and bore into the fruit in order to feed on the pulp (Fonseca, 1934). Once they have penetrated the fruit, larval control becomes impracticable and the fruit is rendered unfit for consumption (Bento et al., 2001).\n\nIn the 1980s, indiscriminate use of insecticides, especially pyrethroids, against a wide variety of pest insects and mites in citrus orchards in Central-Southern Brazil contributed to a drastic reduction of natural enemies, favoring an increase in the population of G. aurantianum (Parra et al., 2004). Starting in the 1990s, yield losses due to the citrus fruit borer were estimated at over US$50 million per year (Anonymous, 2000).\n\nThe sex pheromone of G. aurantianum, (E)-8-dodecenyl acetate and (E)-8-dodecenol, was identified by members of our group in early 2000’s (Leal et al., 2001). At that time, Bento et al. (2001) established strategies for its use in the field, including the number of traps per area, trap positioning on trees, pheromone durability, and control level based on number of males collected per week. In November 2001, the Rural Growers Cooperative (Coopercitrus) placed the synthetic pheromone on the market, focusing on citrus growers in the state of São Paulo after an intense campaign to divulge the technology, train extension agronomists, give presentations, and distribute technical bulletins to citrus producers (Parra et al., 2004).\n\nIn this paper, we report a benefit-cost analysis applied to the citrus industry in the period from 2001 to 2013 in the state of São Paulo, Brazil, based on gross revenues (in US$) corresponding to total production (in boxes of oranges) that growers avoided losing by using traps baited with the sex pheromone of the citrus fruit borer G. aurantianum in the monitored areas. We also discuss strategies for pheromone-baited trap use and its efficiency in the management and control of G. aurantianum.\n\n\nMaterials and methods\n\nThe economic analysis covered the period from November 2001 to December 2013. Monetary results were calculated as losses avoided, i.e., the amount of gross revenues (in US$) corresponding to total production (in boxes of oranges) whose loss was prevented by using traps baited with pheromone of the citrus fruit borer G. aurantianum, in the monitored areas in the state of São Paulo, Brazil.\n\nData on the number of citrus trees in the state of São Paulo and their average yield (boxes/tree) were obtained from the Agricultural Economics Institute (IEA) (IEA, 2015). To calculate the average annual price (US$) of sale of one box of oranges (40.8 kg), we used the average monthly price published by the Center for Advanced Studies on Applied Economics (Cepea) (Cepea, 2015), corresponding to the average amounts in US$ paid to citrus growers per box, on credit, in the state of São Paulo, Brazil, including costs of harvesting and shipping, for oranges of the Pera, Natal and Valencia varieties. Monetary variables were updated to values applicable in June 2014, the final month of data used in this report, using the average exchange rate (PTAX) effective on that month as informed by the Central Bank of Brazil (Bacen, 2015). The reference discount rate considered here was the average annual rate of 4% published by the Special System for Settlement and Custody (Selic) of the Central Bank of Brazil for June 2014 (Bacen, 2015), and the nominal data were transformed into real values using the General Price Index – Internal Availability (IGP-DI), published by the Getúlio Vargas Foundation (FGV, 2015).\n\nThe number of traps baited with G. aurantianum pheromone sold between November 2001 and December 2013, as well as their prices (in US$) were obtained from the Coopercitrus, the only entity responsible for their distribution in the entire state of São Paulo, Brazil. Each year (2001–2013), G. aurantianum was monitored during the citrus harvesting season (~ 6 months). According to Bento et al. (2001), the traps have a durability of one month and cover an area of approximately 10 hectares when used for monitoring. Therefore, six traps/year were installed per 10 hectares monitored. According to available data, the citrus fruit borer can cause yield losses of up to 50% per tree (Parra et al., 2004). However, for our calculations, we considered a 5 to 50% range of losses avoided in the period from November 2001 to December 2013. Costs were calculated based on the prices paid for purchase of the traps and the initial amount invested in research to develop the technology, which was US$50,000 (Parra et al., 2004). Costs of labor for trap installation and monitoring, as well as indirect investments, including use of University resources and researchers and product registration expenses, were not taken into account. Benefits were estimated in the form of losses avoided, by calculating the number of boxes produced in a scenario in which the citrus fruit borer is present, i.e., considering the yield losses that would be caused by the pest if no traps had been used. These losses were then monetized, based on the price of a box of oranges. Finally, the benefit-to-cost ratio was calculated based on total present value, considering both the benefits and the estimated costs of monitoring and control of the citrus fruit borer between 2001 and 2013.\n\n\nResults and discussion\n\nTotal losses avoided by using traps baited with sex pheromone of G. aurantianum in the period from 2001 to 2013 ranged from US$132.7 million to US$1.32 billion in gross revenues. In other words, this was the estimated aggregate total of gross revenues from the sale of oranges that growers avoided losing by using pheromone-baited traps, considering a 5–50% range of potential losses caused by citrus fruit borer infestation in citrus orchards in the state of São Paulo (Table 1; Figure 1). Of note, it is already known that the citrus fruit borer can cause yield losses of up to 50% per tree (Parra et al., 2004)\n\n*Start sales (Nov., 2001)\n\n** 1 box = 40.8 Kg\n\n*** 1 trap/10ha/month, during 6 months (see Bento et al., 2001)(Trap sold per year/6 × 10ha)\n\n**** Area covered by traps (ha)/year × Trees (units/ha/year)\n\nCalculations considered yield loss ranging from a very conservative (5%) up to high (50%) estimates (Parra et al., 2004).\n\nThe total cost of trap purchases from 2001 to 2013 was US$5,065,807.81 (US$5.06 million). It should be noted that some costs were not measured in this study, such as labor costs of the inspections that used to be performed before the traps became available, and the fact that insecticide spraying was once triggered by a 3–5% yield loss caused by the caterpillars of G. aurantianum (Gravena, 1998). Therefore, economic losses due to infested fruit were already occurring in the field, as were expenditures on chemical controls (labor, products and machine time) that were extensively used in the entire area of the orchard. Pheromone-baited traps lowered the costs of inspections (labor) in the entire orchard, in addition to reducing the costs of machine operation and insecticide use, as chemical control became targeted only at areas effectively infested with the insect at quantities above the control level. According to Bento et al. (2001), the use of pheromone-baited traps was shown to be efficient because it monitors adults at their mating stage, enabling growers to apply chemical control before oviposition and subsequent damage to fruits. The authors also showed that, when a control level of six or more males/week was adopted, the average percentage of damaged fruits was 0.6% in the monitored areas. In addition, after successive years of trap use, growers achieved a reduction of approximately 50% in insecticide use to control the citrus fruit borer G. aurantianum (Parra et al., 2004).\n\nThe initial investment in the research that resulted in the development of pheromone-baited trap was US$ 50,000. Therefore, in terms of the governmental costs, the benefit-cost ratio of the initial investment (present value of losses avoided/total investment) ranged from US$ 2,655 to 26,548 per dollar spent with a yield loss of 5–50%, respectively (Figure 2a). In terms of the return for the producer, in which the cost of the traps is included (US$ 5.06 million), the benefit-cost ratio was US$ 12.02 to 120.19 per dollar spent considering yield losses of 5–50% (Figure 2b). These potential losses were based on an estimation of infestation by G. aurantianum in citrus orchards in the state of São Paulo.\n\nGovernmental (A) and producers (B) benefit-to-cost ratio (US$) by investment in research or implementation of pheromone-baited traps to monitor populations of the citrus fruit borer Gymnandrosoma aurantianum between 2001 and 2013 in the state of São Paulo, Brazil and rationalize insecticide sprays.\n\nExcept for the year 2001, when the sex pheromone of G. aurantianum only became available on the market in November, the area monitored during the 12 subsequent years (2002–2013) ranged from 56,600 to 79,100 hectares of citrus (20.4 to 29.4 million trees), corresponding to 9.7 to 13.5% of the area planted with citrus in the state of São Paulo, the main producing region in Brazil.\n\nThese findings reveal a regularity in the sale and use of pheromone-baited traps by citrus growers during that period (2002–2013). Trap sales were relatively stable in that period, with 38,166 units sold per year on average, ranging from 31,970 units (2010) to 47,436 units (2007), possibly due to fluctuations in international prices of orange juice, the main product exported by the Brazilian citrus industry. This regularity suggests a good level of acceptance and application of the technology by growers, and certainly a benefit obtained from its use.\n\nIt worth mentioning that, according to Parra et al. (2004), the total volume of insecticide sprayed in the monitored areas fell by at least 50%. This can possibly be explained by the fact that spraying was only performed in areas (~10 ha) where the pest was found at levels exceeding the damage level thus preserving the other areas and, consequently, the natural enemies within them. In summary, the use of pheromone in traps for monitoring populations of the citrus fruit borer in 12 years led to tangible benefits to growers and priceless environmental savings.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for Figure 1 and Figure 2, 10.5256/f1000research.9195.d129239 (Bento et al., 2016).",
"appendix": "Author contributions\n\n\n\nJMSB, JRPP, EFV, and WSL designed the research. SHGM and ACOA performed cost-benefit analysis. JMSB, SHGM, ACOA, and WSL analyzed the data. JMSB and WSL wrote the manuscript. All authors have agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no conflict of interest.\n\n\nGrant information\n\nSupported in part by São Paulo Research Foundation (FAPESP), Citriculture Defense Fund (Fundecitrus), National Council for Scientific and Technological Development (CNPq) and Coordinating Agency for the Improvement of Higher Education Personnel (CAPES).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank the citrus producers for making their areas available for research and the Rural Growers Cooperative (Coopercitrus) of São Paulo for distributing and advertising the product.\n\n\nReferences\n\nAnonymous: Tecnologia contra o bicho-furão. Revi Fundecitrus. 2000; 96: 8–10.\n\nBacen – Banco Central do Brasil: Câmbio e Capitais Internacionais. Viewed on: 17 July 2015. Reference Source\n\nBento JMS, Parra JRP, de Miranda SHG, et al.: Dataset 1 in: How much is a pheromone worth? F1000Research. 2016. Data Source\n\nBento JM, Vilela EF, Parra JR, et al.: Monitoramento do bicho-furão-dos-citros com feromônio sexual: bases comportamentais para utilização dessa nova estratégia. Laranja. 2001; 22(2): 351–366. Reference Source\n\nBradley SJ, Walker JTS, Wearing CH, et al.: The use of pheromone traps for leafroller action thresholds in pipfruit. Proceedings of the 51st N.Z. Plant Protection. Conference, 1998; 173–178. Reference Source\n\nButenandt A, Beckmann R, Stamm D, et al.: Über den Sexual-Lockstoff des Seidensspinners Bombyx mori. Reindarstellung und Konstitution. Z. Naturforsch. 1959; 14b: 283–284. Reference Source\n\nCameron PJ, Walker GP, Herman TJ, et al.: Development of economic thresholds and monitoring systems for Helicoverpa armigera (Lepidoptera: Noctuidae) in tomatoes. J Econ Entomol. 2001; 94(5): 1104–1112. PubMed Abstract | Publisher Full Text\n\nCepea – Centro de Estudos Avançados em Economia Aplicada – CEPEA. Citros. Viewed on: 25 November 2015. Reference Source\n\nda Fonseca JP: Combate à lagarta das laranjas Gymnandrosoma aurantianum Costa Lima. Chac Quint. 1934; 50: 215–216.\n\nDunkleblum E: Scale insects. J. Hardie, A.L. Minks (Eds.), Pheromones of Non-Lepidopteran Insects Associated with Agricultural Plants. CABI Publishing, Wallingford, UK, 1999; 251–276. Reference Source\n\nEl-Sayed AM, Suckling DM, Wearing CH, et al.: Potential of mass trapping for long-term pest management and eradication of invasive species. J Econ Entomol. 2006; 99(5): 1550–1564. PubMed Abstract | Publisher Full Text\n\nFGV – Fundação Getúlio Vargas. FGV dados. Viewed on: 25 November 2015. Reference Source\n\nGarcia MS, Parra JR: Comparação de dietas artificiais, com fontes proteicas variáveis, para a criação de Ecdytolopha aurantiana (Lima) (Lepidoptera: Tortricidae). Anais da Sociedade Entomológica do Brasil. 1999; 28: 219–232. Publisher Full Text\n\nGravena S: Manejo ecológico de pragas dos citros – aspectos práticos. Laranja. 1998; 19: 61–77.\n\nHowse PE, Stevens ID, Jones OT: Insect Pheromones and Their Use in Pest Management. Chapman and Hall, London. 1998. Publisher Full Text\n\nIBGE – Instituto Brasileiro de Geografia e Estatística: SIDRA: Banco de dados agregados. Viewed on: 25 November 2015. Reference Source\n\nIEA – Instituto de Economia Agrícola: Banco de dados. Viewed on: 25 November 2015. Reference Source\n\nIoriatti C, Anfora G, Tasin M, et al.: Chemical ecology and management of Lobesia botrana. (Lepidoptera: Tortricidae). J Econ Entomol. 2011; 104(4): 1125–1137. PubMed Abstract | Publisher Full Text\n\nJutsum AR, Gordon RFS: Pheromones: importance to insects and role in pest management. Insect pheromone in plant protection. AR Jutsum and RFS Gordan (eds). John Wiley & Sons Ltd UK, 1989; 1–13.\n\nKnight AL, Hull LA: Use of Sex Pheromone Traps to Monitor Azinphosmethyl Resistance in Tufted Apple Bud Moth (Lepidoptera: Tortricidae). J Econ Entomol. 1989; 82: 1019–1026. Publisher Full Text\n\nKrishnaiah K, Zainalabeuddin S, Ganeswara RA, et al.: Pheromone monitoring systems of rice yellow stem borer, Scirpophaga incertulas Walker. Indian Journal of Plant Protection. 1998; 26(2): 99–106. Reference Source\n\nLaurent P, Frérot B: Monitoring of European corn borer with pheromone-baited traps: review of trapping system basics and remaining problems. J Econ Entomol. 2007; 100(6): 1797–1807. PubMed Abstract | Publisher Full Text\n\nLeal WS, Bento JM, Murata Y, et al.: Identification, synthesis, and field evaluation of the sex pheromone of the citrus fruit borer Ecdytolopha aurantiana. J Chem Ecol. 2001; 27(10): 2041–2051. PubMed Abstract | Publisher Full Text\n\nLiebhold AM, Tobin PC: Population ecology of insect invasions and their management. Annu Rev Entomol. 2008; 53: 387–408. PubMed Abstract | Publisher Full Text\n\nMadsen HF, Vakenti JM: Codling Moth: Use of Codlemone®-Baited Traps and Visual Detection of Entries to Determine Need of Sprays. Environ Entomol. 1973; 2: 677–680. Publisher Full Text\n\nMcBrien HL, Judd GJ, Borden JH: Campylomma verbasci (Heteroptera: Miridae): Pheromone-based seasonal flight patterns and prediction of nymphal densities in apple orchards. J Econ Entomol. 1994; 87(5): 1224–1229. Publisher Full Text\n\nParra JRP, Bento JMS, Garcia MS, et al.: Development of a control alternative for the citrus fruit borer, Ecdytolopha aurantiana (Lepidoptera, Tortricidae): from basic research to the grower. Rev Bras Entomol. 2004; 48(4): 561–567. Publisher Full Text\n\nQureshi ZA, Bughio AR, Siddiqui QH, et al.: Seasonal population fluctuation of pink bollworm, Pectinophora gossypiella. (Saund.) (Lep., Gelechiidae) as monitored by gossyplure. J Appl Entomol. 1984; 98(1–5): 43–46. Publisher Full Text\n\nRidgway RL, Silverstein RM, Inscoe MN: Behavior-modifying Chemicals for Insect Management: Applications of Pheromones and Other Attractants. Marcel Dekker, New York. 1990. Reference Source\n\nRothschild GHL, Vickers RA: Biology, ecology and control of the Oriental fruit moth. In Tortricid Pests Their Biology, Natural Enemies and Control (L. P. S. Van Der Geest and H. H. Evenhuis eds.). Elsevier, Amsterdam, Netherlands, 1991; 389–412.\n\nSilverstein RM: Pheromones: Background and potential for use in insect pest control. Science. 1981; 213(4514): 1326–1332. PubMed Abstract | Publisher Full Text\n\nSingh KN, Sachan GC: Asessment of the use of sex pheromone traps in the management of Spodoptera litura F. Indian Journal of Plant Protection. 1993; 21(1): 7–13. Reference Source\n\nThomson DR, Gut LJ, Jenkins JW: Pheromones for insect control: Strategies and successes. In: Hall, F.R., Menn, J.J. (Eds.), Methods in Biotechnology, Biopesticides: Use and Delivery. Humana Press, Totowa, NJ. 1999; 5: 385–412. Publisher Full Text\n\nUSDA – United States Department of Agriculture: Citrus: World Markets and Trade. Viewed on: 20 November 2015. Reference Source\n\nWall C, Garthwaite DG, Smyth JB, et al.: The efficacy of sex-attractant monitoring for the pea moth, Cydia nigricana, in England, 1980–1985. Annals of Applied Biology. 1987; 110(2): 223–229. Publisher Full Text\n\nWitzgall P, Kirsch P, Cork A: Sex pheromones and their impact on pest management. J Chem Ecol. 2010; 36(1): 80–100. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "15137",
"date": "25 Jul 2016",
"name": "A. Cameron Oehlschlager",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper analyzes the monetary benefits accruing to citrus growers adopting a 1 trap / 10 ha pheromone trap monitoring program for citrus fruit borer Gymnamdrosoma aurantianum between 2001 and 2013. The principle savings was in a 50% reduction in insecticide application accompanied by significantly lower fruit damage. This is the first time the economic analysis of long term pheromone trap monitoring to better time insecticide application has been conducted. Because manpower costs of monitoring and spraying are not included, and because spraying uses much more manpower than monitoring, the actual additional income to citrus growers using the monitoring program is very likely more than estimated in this paper. The information in this paper is of significant interest to managers of agricultural enterprises.",
"responses": []
},
{
"id": "15333",
"date": "01 Aug 2016",
"name": "Baldwyn Torto",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWe considered the following issues in our review of the manuscript:\nEconomic analysis was based on direct gross revenue corresponding to total production loss avoided. How was this directly or indirectly attributed to sex pheromones?\nDirect economic benefits are ascribed to costs avoided and cost-benefit ratio per dollar spent on research. What were direct and indirect research costs? How were these measured?\nThe accuracy of monetary values is in part, but importantly, dependent on the accurate treatment of price data. On the pivotal question of price, the reviewer considered the following questions:\n-\n\nIs the average price decomposed into the different contributing variables? -\n\nWhat were the relative contributions of each monetary variable to the aggregated components? - What was the price elasticity? -\n\nto what extent did the assigned exchange rate affect the overall pricing? -\n\nhow stable was the exchange rate over the entire reference period?\n\nAdditional benefits: environmental and ecosystem services?\nReviewers’ specific Comments:\nThe paper is well-written and significant because of the global renewed focus on alternative environmentally-sustainable IPM strategies using non-polluting, non toxic nature-inspired synthetic products, reducing the use of insecticides (and thereby the carbon footprint of their production) in crop and environmental protection.\n\nThe research methodology is adequate. The sample size is adequately representative (9.7% - 13.5% of the total area planted with citrus in Sao Paulo).\n\nThe use of an action level data based on capture with pheromone-baited traps is sufficiently documented/ well exemplified.\n\nThe main thrust of the paper data analysis is based on benefit-cost analysis applied to the citrus industry in the period from 2001 to 2013 in the state of São Paulo, based on gross revenues (in US$) corresponding to total production (in boxes of oranges) that growers avoided losing by using traps baited with the sex pheromone of the citrus fruit borer G. aurantianum in the monitored areas. Whereas the economic value of the pheromone is adequately demonstrated, and the treatment of data, while generally sound, it could be further enhanced by the following considerations:\n4.1. In the calculation of the benefit-cost ratio of the initial investment, i.e. present value of losses avoided/total investment, the reviewer’s considered opinion is that the prices paid for purchase of the traps and the initial amount invested in research to develop the technology (said to be US$50,000) may not adequately define total costs. The authors correctly mention they excluded labor and indirect intellectual investments – which could be considerable. The cost of technology delivery apart from initial research costs is often quite considerable. The authors should make an attempt to estimate these, although understandably these are a complex mix of quantifiable and qualitative values that are not easy to calibrate accurately.\n4.2. In terms of costs avoided, it would also be instructive to attempt to compute all the previous direct and ancillary costs e.g. labor costs of the inspections that used to be performed before the traps became available, in addition to the directly attributed (avoided) costs of previous chemical control.\n4.3. Data on annual average yield is adequately treated. However, price calculations are based on the average monthly price published by the Center for Advanced Studies on Applied Economics (Cepea) (Cepea, 2015), corresponding to the average amounts in US$ paid to citrus growers per box, on credit. In this calculation, price stability is assumed. The authors need to mention if they took account or how they treated supply Vs demand pressure on price, or any subjectively negotiated discounts on price?\nThe accuracy of monetary values is in part, but importantly, dependent on the accurate treatment of price data. On the pivotal question of price, the reviewer considered the following questions:\n- Is the average price decomposed into the different contributing variables? - What were the relative contributions of each monetary variable to the aggregated components? - What was the price elasticity? - to what extent did the assigned exchange rate affected the overall pricing? - how stable was the exchange rate over the entire reference period?\n4.4. In the overall analysis it would be good to capture the computation of price in an equation that shows the relative contribution of each monetary variable to the aggregated components since the monetary variables were derived from multiple independent (assumed non-synchronized) sources e.g. cepea), PTAX), etc.\n4.5 Secondly, in transforming nominal price data into real values using the General Price Index – Internal Availability (IGP-DI) the authors need to discuss the price elasticity and sensitivity to each contributing variable.\n4.6 Thirdly, still on the question of pricing, monetary variables were updated to values applicable in June 2014, and the final month’s PITAX exchange rate used. The authors need to demonstrate to what extent did the assigned exchange rate affected the overall pricing, and explain how stable the exchange rate had been for the entire reference period.\n\nThe authors have made an excellent attempt to compute the actual economic (transactional) value of the the Gymnandrosoma aurantianum sex pheromone. Globally, in the determination of the broader public goods the overall discussion of “how much is a pheromone worth?” could be further enriched if the authors also discussed in greater detail its:\na) Functional value? In terms of environmental ecosystem services provided in addition to economic value?\nb) Evolutionary value of sex pheromones? Maintenance of ecological balance though managed species populations?\nConclusion: The paper is sufficient for indexation, with minor revisions addressing the concerns above.",
"responses": [
{
"c_id": "2148",
"date": "22 Aug 2016",
"name": "Walter Leal",
"role": "Author Response",
"response": "First of all, the authors would like to thank the reviewers for their time and effort to evaluate this submission. We are delighted to hear that the reviewers approved the publication. We appreciate the laudatory comments. Here we address point-by-point some issues raised by Dr. Baldwyn Torto (and Jimmy Pittchar): 4.1. In the calculation of the benefit-cost ratio of the initial investment, i.e. present value of losses avoided/total investment, the reviewer’s considered opinion is that the prices paid for purchase of the traps and the initial amount invested in research to develop the technology (said to be US$50,000) may not adequately define total costs. The authors correctly mention they excluded labor and indirect intellectual investments – which could be considerable. The cost of technology delivery apart from initial research costs is often quite considerable. The authors should make an attempt to estimate these, although understandably these are a complex mix of quantifiable and qualitative values that are not easy to calibrate accurately. RESPONSE: We agree that the initial amount invested in research to develop the technology does not necessarily represent the total costs of technology delivery. As the reviewer mentioned, these estimates are a complex mix of quantifiable and qualitative values that are difficult to be accurately calibrated. Future work will be required to estimate these values for a more precise adjustment. 4.2. In terms of costs avoided, it would also be instructive to attempt to compute all the previous direct and ancillary costs e.g. labor costs of the inspections that used to be performed before the traps became available, in addition to the directly attributed (avoided) costs of previous chemical control. RESPONSE: This is an interesting possibility and we agree with that, but unfortunately we did not recordthese data during the study period. Given the lack of reliable data we prefer not to speculate on these costs and rather focus on effective data. 4.3. Data on annual average yield is adequately treated. However, price calculations are based on the average monthly price published by the Center for Advanced Studies on Applied Economics (Cepea) (Cepea, 2015), corresponding to the average amounts in US$ paid to citrus growers per box, on credit. In this calculation, price stability is assumed. The authors need to mention if they took account or how they treated supply Vs demand pressure on price, or any subjectively negotiated discounts on price? RESPONSE: The prices used in the analysis were the market mean values obtained throughout the study years, reflecting the supply and demand conditions for the product in each year (ie., we did not calculate an overall average for all the years, but rather the annual average of the prices in the market). This was possible because we performed an ‘ex-post’ analysis and already knew the market equilibrium price for each year. The accuracy of monetary values is in part, but importantly, dependent on the accurate treatment of price data. On the pivotal question of price, the reviewer considered the following questions: - Is the average price decomposed into the different contributing variables? RESPONSE: The prices have not been estimated by econometric models. So, explanatory variables were not used to explain the prices behaviour. We worked with historical data to estimate the annual mean prices. - What were the relative contributions of each monetary variable to the aggregated components? RESPONSE: Please, see the previous answer. - What was the price elasticity? RESPONSE: It has not been calculated. - to what extent did the assigned exchange rate affected the overall pricing? RESPONSE: The exchange rate was used to transform the Brazilian currency value (Real) into US currency (USD). We believe that the exchange rate used in this study did not affect directly the overall pricing. - how stable was the exchange rate over the entire reference period? RESPONSE: The exchange rate was used to obtain the value in US dollars (USD). We did not search the exchange rate for the study period or assessed the sensitivity exchange rate facing its variation. 4.4. In the overall analysis it would be good to capture the computation of price in an equation that shows the relative contribution of each monetary variable to the aggregated components since the monetary variables were derived from multiple independent (assumed non-synchronized) sources e.g. cepea), PTAX), etc. RESPONSE: We used observed market equilibrium prices, which were not based on econometric models. 4.5 Secondly, in transforming nominal price data into real values using the General Price Index – Internal Availability (IGP-DI) the authors need to discuss the price elasticity and sensitivity to each contributing variable. RESPONSE: Considering the fact that Brazilian inflation rates are considered high, it is necessary to deflate them, because compared values can be compromised, i.e., it is necessary to remove inflation effect from the period in order to make the prices comparable. For low inflation countries, this methodology is unnecessary. 4.6 Thirdly, still on the question of pricing, monetary variables were updated to values applicable in June 2014, and the final month’s PITAX exchange rate used. The authors need to demonstrate to what extent did the assigned exchange rate affected the overall pricing, and explain how stable the exchange rate had been for the entire reference period. RESPONSE: The exchange rate was used to obtain the value in US dollars (USD). We did not search the exchange rate for the study period or assessed the sensitivity exchange rate facing its variation. 5. The authors have made an excellent attempt to compute the actual economic (transactional) value of the the Gymnandrosoma aurantianum sex pheromone. Globally, in the determination of the broader public goods the overall discussion of “how much is a pheromone worth?” could be further enriched if the authors also discussed in greater detail its: a) Functional value? In terms of environmental ecosystem services provided in addition to economic value? RESPONSE: We agree that the functional value could add important information to our study. However, we consider difficult to measure economically these values in terms of 'ecosystem services'. b) Evolutionary value of sex pheromones? Maintenance of ecological balance though managed species populations? RESPONSE: We thank the referee for this insight. We agree, and believe that future studies should incorporate this relevant perspective."
},
{
"c_id": "2152",
"date": "22 Aug 2016",
"name": "Baldwyn Torto",
"role": "Reviewer Response",
"response": "I think the authors have done a very good job responding to all the queries we raised in our review of the article. Overall, it's a well researched study and well written."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1763
|
https://f1000research.com/articles/5-1396/v1
|
16 Jun 16
|
{
"type": "Software Tool Article",
"title": "search.bioPreprint: a discovery tool for cutting edge, preprint biomedical research articles",
"authors": [
"Carrie L. Iwema",
"John LaDue",
"Angela Zack",
"Ansuman Chattopadhyay",
"Carrie L. Iwema",
"John LaDue",
"Angela Zack"
],
"abstract": "The time it takes for a completed manuscript to be published traditionally can be extremely lengthy. Article publication delay, which occurs in part due to constraints associated with peer review, can prevent the timely dissemination of critical and actionable data associated with new information on rare diseases or developing health concerns such as Zika virus. Preprint servers are open access online repositories housing preprint research articles that enable authors (1) to make their research immediately and freely available and (2) to receive commentary and peer review prior to journal submission. There is a growing movement of preprint advocates aiming to change the current journal publication and peer review system, proposing that preprints catalyze biomedical discovery, support career advancement, and improve scientific communication. While the number of articles submitted to and hosted by preprint servers are gradually increasing, there has been no simple way to identify biomedical research published in a preprint format, as they are not typically indexed and are only discoverable by directly searching the specific preprint server websites. To address this issue, we created a search engine that quickly compiles preprints from disparate host repositories and provides a one-stop search solution. Additionally, we developed a web application that bolsters the discovery of preprints by enabling each and every word or phrase appearing to with articles from preprint servers. This tool, search.bioPreprint, is publicly available at http://www.hsls.pitt.edu/resources/preprint.",
"keywords": [
"bookmarklet",
"grey literature",
"open science",
"peer review",
"pre-print",
"pre-publish",
"search engine",
"server"
],
"content": "Introduction\n\nPreprint servers are online repositories that manage access to manuscripts that have not yet been peer-reviewed or formally published in a traditional manner. Preprint manuscripts are not copyedited, but they do undergo a basic screening process to check against plagiarism, offensiveness, and non-scientific content. Authors may make revisions at any point, but all versions remain available online. It should be noted that the term “preprint” in this context refers to manuscripts posted by the authors themselves onto specific online servers, not articles made available online by publishers a few weeks ahead of traditional publication.\n\nPreprint articles can be more difficult to discover than those published traditionally, as they are not currently indexed in Medline and therefore do not appear in PubMed search results. This suggests that many timely and relevant research reports potentially fall through the cracks, as the time it takes to traditionally publish a biomedical manuscript can take anywhere from a few months to a few years. This lengthy process is seen by researchers to be a hindrance to scientific advancement. In response, there is a developing movement of preprint advocates who propose that preprints play a role in “catalyzing scientific discovery, facilitating career advancement, and improving the culture of communication within the biology community”1. Preprint servers “enable authors to make their findings immediately available to the scientific community and receive feedback on draft manuscripts before they are submitted to journals”2.\n\nThe history, rationale, and controversy surrounding preprint servers and the pace of the current publication process has been well addressed in other manuscripts3–14, news items15–22, and blogs or white papers23–32. We do not intend to duplicate this information here, but suggest exploration of our reference list for an overview of the current state of the topic.\n\nThere are currently only a small number of preprint servers catering to biological and biomedical research manuscripts.\n\narXiv is a venerable preprint server covering physics, mathematics, computer science, nonlinear sciences, statistics, and quantitative biology since 1991. arXiv is funded by Cornell University Library, the Simons Foundation, and many member institutions.\n\nbioRxiv, operated by Cold Spring Harbor Laboratory, covers new, confirmatory, and contradictory results in research ranging from animal behavior and cognition to clinical trials, neuroscience to zoology.\n\nF1000Research, a member of the Science Navigation Group, provides an open science platform for the immediate publication of scientific communication. Posters and slides receive a digital object identifier and are instantly citable. Articles with associated source data are published within a week and made available for open peer review and user commenting. Articles that pass peer review are then indexed in PubMed, Scopus, and Google Scholar. It should be noted that F1000Research is not technically a preprint server, but is included here because it does provide access to articles prior to and during the peer review process. See the Limitations section for details.\n\nPeerJ Preprints covers biological, medical, life, and computer sciences. Their aim is to reduce publishing costs while still efficiently publishing innovative research, with an emphasis on not yet peer-reviewed articles, abstracts, or posters. Submissions are free, can be a draft, incomplete, or final version, and are typically online within a day after editorial approval.\n\nOur intention is to present a resource that facilitates the quick and easy identification and access of scientific content located on preprint servers. The Health Sciences Library System at the University of Pittsburgh (HSLS) developed a tool to help researchers to quickly search preprint databases and discover cutting edge, yet-to-be published or reviewed biomedical research articles, search.bioPreprint (Figure 1). This search engine encompasses a federated search of arXiv, bioRxiv, F1000Research, and PeerJ Preprints. For ease of reading we will continue to refer to all sources of preprint articles as “preprint servers,” including the open science publishing platform F1000Research. We chose to publish this article in F1000Research and bioRxiv in order to support the preprint movement and to elicit feedback on usage of the tool, which will be updated as needed.\n\n\nImplementation\n\nsearch.bioPreprint was created using the proprietary software IBM Watson Explorer, formerly Vivisimo Velocity, version 8.0-2 (IBM Corp, Armonk, New York, USA) to generate a meta search engine that compiles search term results from a pre-selected list of multiple sources into a single list ordered by the relevance of matching query terms. The results can then be further filtered by Source (e.g., the preprint servers of origin) or by Topic (e.g., microcephaly for a Zika virus search). The Topic search is accomplished via clustering, meaning the search results are organized on the fly by similarity in subject matter. Additionally, a “remix” link displayed next to the clustered topics reveals new secondary topics. This is done by clustering the same search results again, but explicitly ignoring the topics that were used in the initial clustering process.\n\nThe Health Sciences Library System at the University of Pittsburgh has repeatedly utilized IBM Watson Explorer software to develop, implement, and maintain several federated search engines focused on a variety of topics. These include: search.HSLS.OBRC –a portal for discovering bioinformatics databases and software via the Online Bioinformatics Resource Collection33, Clinical Focus –a portal providing quick access to high-quality clinical information34, and Clinical eCompanion –a portal with information for primary care35. Similarly, the U.S. National Library of Medicine (NLM) utilized the same software to create search engines for MedlinePlus, MedlinePlus en Español, and the NLM library website.\n\nThe search engine was created following the software manufacturer’s protocol. Briefly, the search url and parameters are entered for each site, then the results are selected based on the XPath of the results within the HTML page. Finally, each individual source is bundled into a single source to provide one search for multiple sites.A maximum of 200 total results are returned based on the licensing agreement with IBM; this also contributes to a short wait for return of results. The selected sources for retrieving preprint articles using search.bioPreprint are: (1) the quantitative biology section of arXiv.org, (2) bioRxiv, (3) F1000Research and (4) PeerJ Preprints.\n\nAs an example, typing a single-word query term, such as CRISPR, into the search box results in ninety-one preprint articles culled from the aformentioned preprint servers (Figure 2, searched on 2 May 2016). Clicking on an article title redirects to that article at its original source. Search results may be narrowed by Topic or Source using the filters on the left side of the page. Using the CRISPR example, the ninety-one search results are grouped into shared Topics: fourteen articles on “Bacterial,” twelve articles on “Protein,” six articles on “Genome engineering,” etc. Expanding individual topics reveals a list of subtopics: clicking on the topic “Protein” redistributes the twelve articles into subtopics, including “CRISPR-Cas9,” “Image, Palindromic Repeat,” “Mutants, Generated,” etc. Clicking on a topic or subtopic reconfigures the search results to limit to these filtered articles.\n\nAt left is the default view by Topic. (2 May 2016).\n\nClicking on the “remix” button appearing next to “Top 91 results” regroups the original search results into additional topics such as “Cells,” “Advances,” “Drosophila,” etc that are not present in the first results iteration (Figure 3). This provides another opportunity to discover pertinent preprint articles, especially if a large number of results is returned.\n\n(2 May 2016).\n\nThe search results may also be filtered by Source. Selecting this will change the default display of topic-focused clusters to articles organized by Source, which in the current iteration is one of the four preprint servers searched by this tool: nineteen from F1000Research, two from PeerJ Preprints, six from arXiv, and sixty-five from bioRxiv (Figure 4).\n\n(2 May 2016).\n\nQuotation marks are recommended for searches with exact phrases, e.g., Zika virus. The necessity of this was discovered after examing the search parameters of the various preprint servers. As one of the preprint servers by default joins words in a multi-word query with the Boolean operator “OR” then a search for a phrase such as zika virus produces multiple articles where the only matching term is virus. Using quotation marks for a search of more than one word mitigates this problem and considerably improves the quality of results. A search for “zika virus” thus produces seventy-nine articles that are topically filtered into “Zika virus infection,” “Microcephaly,” “Discovery,” “Dengue Virus,” etc (searched on 2 May 2016).\n\nThe “Search within clusters” box allows for searching within the search results, and can be used to identify specific articles within the cohort of Zika virus preprints that are not immediately apparent from topical clustering. Entering vaccine in the search box highlights the topics and subtopics containing articles bearing the word vaccine: under “Zika virus infection” is “Preventing Zika Virus Infection;” under Dengue Virus is “Antibodies, Vaccine” and “Community, Vector.” Selection of highlighted topics or subtopics reconfigures the results to limit to vaccine-related Zika virus preprints (Figure 5).\n\n(2 May 2016).\n\nA bookmarklet is a special type of web browser widget containing an embedded software command that extends the application of the browser by adding a one-click function as a bookmark. We created a bioPreprint-bookmarklet using JavaScript in order to seamlessly integrate a search for any word or phrase from any web page with the information stored in preprint servers. After dragging/dropping the bioPreprint-bookmarklet into any web browser, the next step is to highlight a word or phrase of interest then click the bookmarklet. This will result in a pop-up window displaying preprint articles containing the text of interest (Figure 6).\n\n(2 May 2016).\n\nAll web browsers that support JavaScript (Google Chrome, Mozilla FireFox, Internet Explorer, Apple Safari, Opera) are compatible with the bookmarklet. In case the favorites/bookmark bar is not visible we provide instructions for displaying it on commonly used browsers. A video describing how to install the bookmarklet in a web browser is also available.\n\n\nUse cases\n\nImagine a researcher is searching PubMed for articles on “RNA-seq quantification” and comes across a paper recently published in Nature Biotechnology, “Near-optimal probabilistic RNA-seq quantification”36. This paper introduces a new software program, Kallisto, that analyzes RNA-seq data by two orders of magnitude faster than previously used software. This is notable as it removes the computational bottleneck for RNA-seq data analysis. After reading about this new software, the researcher decides to check whether its been widedly adopted by perusing the published literature.\n\nA search in PubMed with the search term “Kallisto” results in only the original article (searched on 2 May, 2016). This is well within expectations, considering the recent publication date of the article, 4 April 2016. There has not been enough time for researchers to know about the software, let alone write papers citing it.\n\nTo continue to try and gauge the usage of Kallisto in RNA-seq data analysis, the researcher might take an alternative approach: instead of searching PubMed, try searching for preprint articles. This can be achieved with a single click of the bioPreprint-bookmarklet once it is installed in the researcher’s web browser. Upon viewing the article abstract on the PubMed search results page, highlighting the word “Kallisto,” and clicking the bioPreprint-bookmarklet, a pop-up appears with the search.bioPreprint search results: sixteen preprint articles, two from arXiv, thirteen from bioRxiv, and one from F1000Research (searched on 2 May, 2016). Interestingly, the second article on the results page is the preprint version of the Nature Biotechnology paper on Kallisto software, submitted to the arXiv preprint server (Figure 6). The authors submitted their preprint on 11 May 2015, almost one year before its publication in Nature Biotechnology, with concomitant indexing by PubMed37.\n\nIt is worth noting that since the availability of the Kallisto paper as a preprint, fifteen preprint articles have cited the use of Kallisto software38–51, searched on 2 May, 2016). These articles cover numerous topics, including development of new software, single cell RNA-seq analysis, and quantification of the relative abundance of transcripts in various experimental settings.\n\nA student gathering information from the internet about the regulation of gene expression happens upon the GTEx Project Community Scientific Meeting website. GTEx stands for the Genotype-Tissue Expression project (GTEx), which aims to develop an atlas of human gene expression and its regulation across various tissue types. Intrigued by the scope of this project, the student is curious to know how GTEx project data have been utilized in research.\n\nThe bioPreprint search engine and bookmarklet can quickly satisfy the student’s curiosity by providing easy access to GTEx-related articles hosted by various preprint servers that may or may not be published “in print” yet. This process is simple, unique, and the student doesn’t even need to leave the current web page to go on a literature hunt. Rather, all GTEx-related articles will appear in a new window with only two clicks, the first highlighting the word GTEx and the second on the previously-installed bioPreprint-bookmarklet. The result is sixty-seven articles showcasing the use of GTEx data in a variety of research topics including “Genome Wide Association Studies,” “Allele, Specific expression,” “Expression Quantitative Trait Loci,” etc (searched on 2 May 2016).\n\n\nConclusions\n\nThese use cases emphasize the power of the bioPreprint search engine and associated bookmarklet in delivering scientific research articles that are not only hard-to-find and yet-to-be traditionally published, but also on demand at the point of reading. And the “point of reading” can be anything on the web: journal articles, news items, blogs, PubMed/Google Scholar search results, etc.\n\nUntil the creation of search.bioPreprint there has been no simple way to identify biomedical research published in a preprint format, as they are not typically indexed and are only discoverable by directly searching the preprint server websites (articles that pass peer review in F1000Research are the exception). search.bioPreprint is a one-stop-shop for finding these types of articles, and an important contribution to the preprint movement. During the final stages of manuscript preparation an online database aiming to index preprint articles was launched, PrePubMed, which despite appearances is not an official resource from the National Library of Medicine (NLM), the National Center for Biotechnology Information (NCBI), or PubMed. We want to acknowledge this new resource, but emphasize that search.bioPreprint offers not only full text searching, but also topical and source-based clustering of results. In addition, our tool has been available since mid-February 2016, around the same time as the ASAPbio meeting, where it was mentioned during discussions.\n\nThe underlying technology upon which search.bioPreprint was built is flexible enough to integrate additional resources into the search engine. As new preprint servers are introduced, search.bioPreprint will incorporate them and continue to provide a one-stop solution for finding preprint articles. We welcome feedback that introduces new preprint resources and addresses usability concerns.\n\nThe bioPreprint-bookmarklet enables each and every word or phrase appearing on any website to be integrated with information in articles stored in preprint servers. The on-demand delivery of preprint articles at the point of reading enables researchers to discover brand new pre-published articles quickly and be updated with cutting edge, yet-to-be-reviewed information that is challenging to discover by traditional literature searching methods. Our intention is that the combined use of the aforementioned tools helps to fulfill the unmet need of the scientific community for immediate dissemination of research outcomes, ultimately resulting in improved scientific communication and far-ranging insights and innovations.\n\n\nLimitations\n\nWhile arXiv, bioRxiv, and PeerJPreprints are considered to be preprint servers, F1000Research belongs to a separate class. It offers a unique publishing platform in which a transparent peer review process is integrated into the article publication practice and thus holds three categories of articles based on peer review status: (1) recently submitted and awaiting peer review, (2) passed peer review, and (3) not passed by peer reviewers. Only articles that pass the peer review process are indexed in literature databases such as PubMed. F1000Research permanently hosts all articles irrespective of peer review status. Therefore, it represents a blended system of preprint server and traditional online journal. Search.bioPreprint does not separate these three types of F1000Research articles and therefore returns both non-peer reviewed and reviewed articles together in the search results. Nevertheless, the peer review status is easily visible when searchers are directed to the F1000Research site from the search.bioPreprint search results. As F1000Research hosts many articles whose peer review status (before passing peer review) could be considered the equivalent of preprints, we decided to include this as a source of preprint articles. Users should note a key difference, however, as all articles in F1000Research are committed to formal peer review and should therefore not be submitted to any additional journals.\n\nThe quality of the search results generated by the bioPreprint search engine is confined by the search parameters of the individual preprint servers. If the preprint servers alter their search algorithms, a concomitant adjustment of underlying codes used by the bioPreprint search engine is often required. Unfortunately, this can be done without any public notification and is only discoverable upon a thorough analysis of bioPreprint search results. The University of Pittsburgh Health Sciences Library System has a quality check team involving two librarians to ensure the accuracy of search.bioPreprint results. The team routinely compares the search results produced by several preset query terms with the previous results and reports any discrepancies to the development team.\n\nThe average time taken to display search results is not always optimal. The speed of the search.bioPreprint results return stems from multiple factors: individual preprint servers’ searching speed, efficiency of the IBM Watson Explorer software, and computational power of the server hosting the bioPreprint search engine. While some contributing factors are outside of our control, efforts will be undertaken to speed up the search process by continually upgrading the power of the host server.\n\n\nSoftware availability\n\nsearch.bioPreprint is freely accessible at http://www.hsls.pitt.edu/resources/preprint. The preprint search engine was created using the software, IBM Watson Explorer, formerly known as Vivisimo Velocity. IBM Watson Explorer is a proprietary software, hence, its source code is not available.\n\nThe bioPreprint-bookmarklet is freely available at http://hsls.pitt.edu/biopreprint-infobooster.\n\nThe JavaScript code embedded in the bookmarklet is:\n\n“javascript:(function(){(function(t,u,w){t=''+(window.getSelection%3Fwindow.getSelection():document.getS election%3Fdocument.getSelection():document.selection%3Fdocument.selection.createRange().text:'');u=t %3F'http://search.hsls.pitt.edu/vivisimo/cgi-bin/query-meta%3Fv%253Aproject=preprint% 26query=%2522'+encodeURIComponent(t)+'%2522':'';w=window.open(u,'_blank','height=750,width= 700,scrollbars=1');w.focus %26%26 w.focus();if(!t){w.document.write('<html><head><title></title></head><body style=\"padding:1em;font-family:Helvetica,Arial\"><br/><p>First%2C highlight a word or a group of words from any website that you are browsing (journal article%2C PubMed search result%2C news article%2C blog%2C etc.)%2C and then click on this bookmarklet to retrieve cutting edge%2C yet-to-be published or reviewed biomedical research articles related to your selected word(s).</a><p>Check the <a href=\\\"http://media.hsls.pitt.edu/media/BioPreprint_ac0316.mp4\\\">How to Video</a>for instruction.</p><br/><p><img src=\"http://www.hsls.pitt.edu/sites/all/themes/liberry_front/logo.png\" alt=\"HSLS Logo\"></p><script>var q=document.getElementById(\"q\"),v=q.value;q.focus();q.value=\"\";q.value=v;</script></body></html>'); w.document.close();}})()})();”",
"appendix": "Author contributions\n\n\n\nAC conceived the concept and wrote the Implementation section. CI created the search.bioPreprint logo, assisted with concept refinement and design, and wrote all documentation, including preparation of the initial drafts of the manuscript. AC and CI created the figures. JL developed the search engine. AZ created the bookmarklet and webpage. All authors were involved in revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors wish to gratefully acknowledge the following individuals for their help with various aspects of the creation of search.bioPreprint and manuscript preparation: Peter Coles for writing a blog on the insightful use of bookmarklets, Julia Dahm for the creation of the video describing how to install the bookmarklet, Melissa Ratajeski for providing helpful comments on the manuscript, Nancy Tannery for providing helpful comments on the manuscript and offering general support for this project, and Fran Yarger for offering general support for this project.\n\n\nReferences\n\nASAPbio: Accelerating Science and Publication in Biology. [cited 2016 Mar 29]. Reference Source\n\nbioRxiv. [cited 2016 Mar 29]. Reference Source\n\nDesjardins-Proulx P, White EP, Adamson JJ, et al.: The case for open preprints in biology. PLoS Biol. 2013; 11(5): e1001563. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFisher D, Parisis N: Social influence and peer review: Why traditional peer review is no longer adapted, and how it should evolve. EMBO Rep. 2015; 16(12): 1588–1591. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFord E: Open peer review at four STEM journals: an observational overview [version 2; referees: 2 approved, 2 approved with reservations]. F1000Res. 2015; 4: 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu C, Zhang Y, Chen G: Exploring a New Model for Preprint Server: A Case Study of CSPO. J Acad Libr. 2010; 36(3): 257–262. Publisher Full Text\n\nHutchins BI, Yuan X, Anderson JM, et al.: Relative Citation Ratio (RCR): A new metric that uses citation rates to measure influence at the article level. bioRxiv. 2015; 029629. Publisher Full Text\n\nLauer MS, Krumholz HM, Topol EJ: Time for a prepublication culture in clinical research? Lancet. 2015; 386(10012): 2447–2449. PubMed Abstract | Publisher Full Text\n\nMcDowell G: Junior biomedical scientists and preprints [version 1; referees: 1 approved with reservations]. F1000Res. 2016; 5: 294. Publisher Full Text\n\nPowell K: Does it take too long to publish research? Nature. 2016; 530(7589): 148–151. PubMed Abstract | Publisher Full Text\n\nTomaiuolo NG, Packer JG: Preprint Servers: Pushing the Envelope of Electronic Scholarly Publishing. Search Medford N J. 2000; 8(9). Reference Source\n\nTracz V, Lawrence R: Towards an open science publishing platform [version 1; referees: 2 approved]. F1000Res. 2016; 5: 130. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVale RD: Accelerating scientific publication in biology. Proc Natl Acad Sci U S A. 2015; 112(44): 13439–13446. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarne V: Rewarding reviewers - sense or sensibility? A Wiley study explained. Learned Publishing. 2016; 29(1): 41–50. Publisher Full Text\n\nCallaway E: Preprints come to life. Nature. 2013; 503(7475): 180. PubMed Abstract | Publisher Full Text\n\nCallaway E: Geneticists eye the potential of arXiv. Nature. 2012; 488(7409): 19. PubMed Abstract | Publisher Full Text\n\nCallaway E, Powell K: Biologists urged to hug a preprint. Nature. 2016; 530(7590): 265. PubMed Abstract | Publisher Full Text\n\nGibney E: Open journals that piggyback on arXiv gather momentum. Nature. 2016; 530(7588): 117–118. PubMed Abstract | Publisher Full Text\n\nHarmon A: Handful of Biologists Went Rogue and Published Directly to Internet. 2016. [cited 2016 Mar 28]. Reference Source\n\nKaiser J: New Preprint Server Aims to Be Biologists’ Answer to Physicists' arXiv. Science. AAAS. 2013. [cited 2016 Mar 29]. Reference Source\n\nPalmer KM: A Rainbow Unicorn Wants to Transform Biology Publishing. WIRED. 2016. [cited 2016 Mar 28]. Reference Source\n\nTaking the online medicine. The Economist. 2016. [cited 2016 Mar 28]. Reference Source\n\nBirney E: 10,000 Up. 2015. [cited 2016 Mar 29]. Reference Source\n\nBrown CT: A good way to publish -- arXiv FTW. 2012. [cited 2016 Mar 29]. Reference Source\n\nCurry S: Combining preprints and post-publication peer review: a new (big) deal? Reciprocal Space. 2016. [cited 2016 Mar 29]. Reference Source\n\nHimmelstein D: The History of Publishing Delays. 2016. [cited 2016 Mar 29]. Reference Source\n\nThomason A: Is Scientific Publishing About to Be Disrupted? ASAPbio, Briefly Explained – The Ticker - Blogs - The Chronicle of Higher Education. 2016. [cited 2016 Mar 28]. Reference Source\n\nWhite E: Which preprint server should I use? Jabberwocky Ecology. The Weecology Blog on WordPress.com. 2014. [cited 2016 Mar 29]. Reference Source\n\nTracz V, Lawrence R: The role of preprints in publishing. ASAPbio. 2016. [cited 2016 Mar 29]. Reference Source\n\nSmith R: A better way to publish science-BMJ Blogs. 2015. [cited 2016 Mar 30]. Reference Source\n\nEisen M, Vosshall LB: Coupling Pre-Prints and Post-Publication Peer Review for Fast, Cheap, Fair, and Effective Science Publishing. ASAPbio. 2016. [cited 2016 Mar 30]. Reference Source\n\nHeard S: Post-publication peer review and the problem of privilege. Scientist Sees Squirrel on WordPress.com. 2015. [cited 2016 Mar 30]. Reference Source\n\nChen YB, Chattopadhyay A, Bergen P, et al.: The Online Bioinformatics Resources Collection at the University of Pittsburgh Health Sciences Library System--a one-stop gateway to online bioinformatics databases and software tools. Nucleic Acids Res. 2007; 35(Database issue): D780–D785. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTannery NH, Epstein BA, Wessel CB, et al.: Impact and user satisfaction of a clinical information portal embedded in an electronic health record. Perspect Health Inf Manag. 2011; 8: 1d. PubMed Abstract | Free Full Text\n\nWessel C, LaDue J, Dahm J: Clinical ECompanion: Development of a Point-of-Care Information Tool. Toronto, Ontario, Canada: Medical Library Association; 2016. Reference Source\n\nBray NL, Pimentel H, Melsted P, et al.: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016; 34(5): 525–7. PubMed Abstract | Publisher Full Text\n\nBray N, Pimentel H, Melsted P, et al.: Near-optimal RNA-Seq quantification. arXiv.org. 2015. Reference Source\n\nArakawa K, Yoshida Y, Tomita M: Genome sequencing of a single tardigrade Hypsibius dujardini individual. bioRxiv. 2016. Publisher Full Text\n\nGibilisco L, Zhou QI, Mahajan S, et al.: The evolution of alternative splicing in Drosophila. bioRxiv. 2016. Publisher Full Text\n\nHavens LA, MacManes MD: Characterizing the Adult and Larval Transcriptome of the Multicolored Asian Lady Beetle, Harmonia axyridis. bioRxiv. 2015. Publisher Full Text\n\nHensman J, Papastamoulis P, Glaus P, et al.: Fast and accurate approximate inference of transcript expression from RNA-seq data. arXiv.org. 2015. Reference Source\n\nKordonowy LL, MacManes MD: Characterization of a Male Reproductive Transcriptome for Peromyscus eremicus (Cactus mouse). bioRxiv. 2016. Publisher Full Text\n\nMacManes MD: Establishing evidenced-based best practice for the de novo assembly and evaluation of transcriptomes from non-model organisms. bioRxiv. 2016. Publisher Full Text\n\nMorgan AP, Holt JM, McMullan RC, et al.: The evolutionary fates of a large segmental duplication in mouse. bioRxiv. 2016. Publisher Full Text\n\nNtranos V, Kamath GM, Zhang J, et al.: Fast and accurate single-cell RNA-Seq analysis by clustering of transcript-compatibility counts. bioRxiv. 2016. Publisher Full Text\n\nPai AA, Baharian G, Sabourin AP, et al.: Widespread shortening of 3’ untranslated regions and increased exon inclusion are evolutionarily conserved features of innate immune responses to infection. bioRxiv. 2016. Publisher Full Text\n\nRzepiela AJ, Vina-Vilaseca A, Breda J, et al.: Exploiting variability of single cells to uncover the in vivo hierarchy of miRNA targets. bioRxiv. 2015. Publisher Full Text\n\nPatro R, Duggal G, Kingsford C: Accurate, fast, and model-aware transcript expression quantification with Salmon. bioRxiv. 2015. Publisher Full Text\n\nSoneson C, Matthes KL, Nowicka M, et al.: Differential transcript usage from RNA-seq data: isoform pre-filtering improves performance of count-based methods. bioRxiv. 2015. Publisher Full Text\n\nSoneson C, Love MI, Robinson MD: Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences [version 2; referees: 2 approved]. F1000Res. 2016; 4: 1521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStubbington MJT, Lönnberg T, Proserpio V, et al.: Simultaneously inferring T cell fate and clonality from single cell transcriptomes. bioRxiv. 2015. Publisher Full Text"
}
|
[
{
"id": "14404",
"date": "04 Jul 2016",
"name": "Prachee Avasthi",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nsearch.bioPreprint is a useful tool for the scientific community and this article nicely outlines the key features and use cases for the service. A particular strength of the tool is the site maintenance and quality control by the University of Pittsburgh Health Sciences Library System. Also, the flexibility to incorporate new preprint servers due to continued site support is a benefit. The development of a bookmarklet for this purpose is novel, convenient and very useful.\n\nI have a few comments for authors to address:\n\nThere is no mention in the abstract that preprints can be found through Google Scholar. While there is a temporary bug associated with Google Scholar’s treatment of subsequently published preprints (documented in the following blog posts by Dr. Wilke), preprints are indeed searchable across platforms through Google Scholar. Please mention this and note any benefits of your tool. Google Scholar is mentioned on p2 but only in the context of indexing peer reviewed publications.\nhttp://serialmentor.com/blog/2014/11/1/the-google-scholar-preprint-bug\nhttp://serialmentor.com/blog/2015/10/8/Google-Scholar-bug-redux\n\nThe search results on search.bioPreprint are ordered by relevance with no option to re-sort. After some time searching using search.bioPreprint, the need to sort results by date became quickly evident. Likely, many users will be interested to know what new preprints are available across platforms since a previous search. A search term on Google Scholar followed by “preprint” and one click on the “sort by date” link produces date-sorted preprints from multiple sources (biorxiv, arrive, peerj etc). Similarly, while prepubmed.org also does not have sort functionality, the default ordering appears to be by date. I am unsure how difficult sort functionality would be to implement but some mention of this issue and any plans to implement such a feature in the future is recommended.\n\nOutside of benefits to readers, an additional benefit I can see to a cross-platform preprint search engine is that it becomes easier for journal editors to identify/solicit submission of preprints and for grant reviewers to find preprints prior to journal publication. While these are also benefits of preprints in general, which is outside the scope of this article, authors may want to include this additional motivation/rationale as these are also particular benefits of a cross-platform tool. For example, a reviewer may be more likely to read a pending publication if they don’t need to search many different sites.\n\nIt seems the mention of prepubmed.org in the text is limited to a discussion of priority. Since search.bioPreprint is sold as a one-stop shop, I would have instead liked to see a concise comparison with the other cross-platform searches (prepubmed and also Google Scholar) to help users clearly identify any feature differences or benefits of using search.Biopreprint. The unique features of search.Biopreprint are described throughout the article but a concise comparison or table of features/search result data would be advantageous to readers.\n\nThis article was posted on both F1000Research and bioRxiv. Interestingly, only the bioRxiv version appears as a search result on both search.bioPreprint and Google Scholar, while prepubmed finds both versions. I am curious if some modification by the quality check team would fix this problem or if there is some inherent limitation of search.bioPreprint for preprints posted on more than one server.\n\nOverall, this article clearly describes usage and features of a new tool for cross-platform preprint search that appears to have the advantages of continuous maintenance, useful topic filtering and associated bookmarklet.",
"responses": [
{
"c_id": "2054",
"date": "05 Jul 2016",
"name": "Jordan Anaya",
"role": "Reader Comment",
"response": "Hi, I am the creator of PrePubMed. In response to your comment, yes, PrePubMed sorts articles by date, specifically the date of the earliest version of the preprint that was indexed. In addition, I would like to comment on search.bioPreprint's ability to perform full text searches. I noticed that the PeerJ Preprints search engine does not perform a full text search, only title and abstracts. As a result, I think the claim of a full text search should be qualified."
},
{
"c_id": "2076",
"date": "20 Jul 2016",
"name": "Ansuman Chattopadhyay",
"role": "Author Response",
"response": "We sincerely thank the reviewer for the constructive review and thoughtful suggestions. Specifically, we appreciate the comments that search.bioPreprint is “a useful tool for the scientific community” and that “the development of a bookmarklet is novel, convenient and very useful.” We submitted a revised version of our paper that addresses the reviewer’s concerns, and we hope the reviewer finds this improved version acceptable without further revision. Reviewer Comment 1: There is no mention in the abstract that preprints can be found through Google Scholar. While there is a temporary bug associated with Google Scholar’s treatment of subsequently published preprints (documented in the following blog posts by Dr. Wilke), preprints are indeed searchable across platforms through Google Scholar. Please mention this and note any benefits of your tool. Google Scholar is mentioned on p2 but only in the context of indexing peer reviewed publications. Author Response: Thanks for showing us the workaround for retrieving preprint articles via Google Scholar (GS) by adding “preprint” as a text with the search term(s). We missed the preprint search capability of GS as it is not listed as a valid filter such as “include patents.” However, we noticed that the articles retrieved by a preprint search using GS are not always valid preprints. For example, a search for “asthma preprint” retrieves many articles already published that were never available as a preprint. When the search results (2,430) are sorted by relevence, the third and fourth citations from the top, “Heliox vs air-oxygen mixtures for the treatment of patients with acute asthma: a systematic Overview” by AMH Ho etal. and “Genomic approaches to understanding Asthma” by LJ Palmer etal., are published articles from Elsevier journal Chest (vol 123, issue 3) and from the journal Genome Research (CSH Press), respectively. After investigating a few non-preprint articles, we found the reason for the inclusion of non-preprint published articles in the GS search result is either a mention of a preprint article in the reference list or in the acknowledgement section. The results for another search,“CRISPR preprint,” results in 472 articles by GS and here also the third citation from the top is a published paper from the Journal Genome Research (CSH Press). We added a few sentences on using GS to search for preprints under Conclusions: “Google Scholar (GS), a popular scholarly literature search engine that provides cross-discipline search functionality, does not include preprint articles as a filter option. Hence, many avid GS users try a workaround by including preprint with the query term, (E.g., “asthma preprint” or “CRISPR preprint”) with the assumption of retrieving only preprint articles fetched from major preprint servers. In contrast, the GS search results in a mixed population of articles comprising both actual preprints and peer-reviewed published articles in which the term “preprint” appears somewhere in the full text of the article.” Reviewer Comment 2: The search results on search.bioPreprint are ordered by relevance with no option to re-sort. After some time searching using search.bioPreprint, the need to sort results by date became quickly evident. Likely, many users will be interested to know what new preprints are available across platforms since a previous search. A search term on Google Scholar followed by “preprint” and one click on the “sort by date” link produces date-sorted preprints from multiple sources (biorxiv, arrive, peerj etc). Similarly, while prepubmed.org also does not have sort functionality, the default ordering appears to be by date. I am unsure how difficult sort functionality would be to implement but some mention of this issue and any plans to implement such a feature in the future is recommended. Author Response: We acknowledge the pressing need for an option to sort preprint articles by date as both reviewers mentioned it. We are actively working to bring this functionality into search.bioPreprint. We want to stress that the “Sort by date” feature offered by Google Scholar (GS) is abysmal. It drastically drops the number of retrieved articles compared to the default search results. For example, a GS search for“Asthma preprint” retrieves 2,430 citations by the default “Sort by relevence” option, but displays only 6 articles after selecting “Sort by date.” The same thing happens for another search: “Crispr preprint” – Sort by relevance=471; Sort by date=5. In the revised manuscript we mention the need for reordering the search results by date by adding a new paragraph at the end of the Limitations section: ”Currently, the search.bioPreprint default search results are ordered by relevance without any option to re-sort by date. The authors are aware of the pressing need for this added feature and if possible will incorporate it into the next version of the search tool.” Reviewer Comment 3: Outside of benefits to readers, an additional benefit I can see to a cross-platform preprint search engine is that it becomes easier for journal editors to identify/solicit submission of preprints and for grant reviewers to find preprints prior to journal publication. While these are also benefits of preprints in general, which is outside the scope of this article, authors may want to include this additional motivation/rationale as these are also particular benefits of a cross-platform tool. For example, a reviewer may be more likely to read a pending publication if they don’t need to search many different sites. Author Response: Thanks for the suggestion. We added a sentence under Conclusion: “Referees during the grant or journal article review process might also find this bookmarklet useful as it quickly retrieves pre-published articles via the cross-platform preprint search.” Reviewer Comment 4: It seems the mention of prepubmed.org in the text is limited to a discussion of priority. Since search.bioPreprint is sold as a one-stop shop, I would have instead liked to see a concise comparison with the other cross-platform searches (prepubmed and also Google Scholar) to help users clearly identify any feature differences or benefits of using search.Biopreprint. The unique features of search.Biopreprint are described throughout the article but a concise comparison or table of features/search result data would be advantageous to readers. Author Response: Thanks for the suggestion. Google Scholar does not provide an option to limit searches to preprint articles, and the workaround of including “preprint” with the query term results in a mixed population of articles comprising both actual preprints and peer-reviewed published articles. We are hesitant to consider Google Scholar as a preprint search engine and comparable to search.bioPreprint and PrePubMed. The purpose of this article is to present search.bioPreprint as a means to locate preprint articles, and its release pre-dates PrePubMed. We leave it to others to determine the pros and cons of using search.bioPreprint, and hope that they leave comments so we can improve the tool when possible. Reviewer Comment 5: This article was posted on both F1000Research and bioRxiv. Interestingly, only the bioRxiv version appears as a search result on both search.bioPreprint and Google Scholar, while prepubmed finds both versions. I am curious if some modification by the quality check team would fix this problem or if there is some inherent limitation of search.bioPreprint for preprints posted on more than one server. Author Response: We appreciate your efforts in discovering this search error. The Health Sciences Library System’s quality check team has investigated this issue and is working on a solution. We anticipate a quick fix of this problem."
},
{
"c_id": "2078",
"date": "14 Jul 2016",
"name": "Ansuman Chattopadhyay",
"role": "Author Response",
"response": "Jordan , We appreciate your careful reading of our article and clarification of PeerJ’s search limitations. In the latest version we revised the text to read “We want to acknowledge this new resource, but emphasize that search.bioPreprint offers not only full text searching (with the exception of PeerJ Preprints), but also topical and source-based clustering of results. In addition, our tool has been available since mid-February 2016, around the same time as the ASAPbio meeting, where it was mentioned during discussions.”"
},
{
"c_id": "2083",
"date": "14 Jul 2016",
"name": "Jordan Anaya",
"role": "Reader Comment",
"response": "Thank you for the change. I also recently noticed that the search engine for arXiv q-bio does not perform a full text search. arXiv recently added an experimental full text search option which is available under \"Advanced Search\": https://arxiv.org/find, however it appears to be limited. Regardless, it seems that search.bioPreprint does not utilize this experimental full text search. Can you comment on your plans for arXiv full text searching? Thanks, Jordan"
},
{
"c_id": "2099",
"date": "21 Jul 2016",
"name": "Jordan Anaya",
"role": "Reader Comment",
"response": "I would like to comment on Google Scholar's ability to search for preprints. As covered by Jessica Polka in this blog post you can use the advanced search and type in 'bioRxiv OR “PeerJ Preprints” OR f1000Research OR arxiv' into the \"published in\" field. Google Scholar also indexes PDFs, so it is possible to perform a full text search for journals such as PeerJ Preprints."
},
{
"c_id": "2725",
"date": "25 May 2017",
"name": "Jordan Anaya",
"role": "Reader Comment",
"response": "I would like to clarify and/or raise some issues with this article and accompanying comments. 1. Reviewers Prachee Avasthi and Cynthia Wolberger both emphasized the importance of being able to sort by date, and in response the article was edited to say: \"Currently, the search.bioPreprint default search results are ordered by relevance without any option to re-sort by date. The authors are aware of the pressing need for this added feature and if possible will incorporate it into the next version of the search tool.\" However, it has been nearly a year and this feature has not been added. 2. The article states: \"Until the creation of search.bioPreprint there has been no simple and efficient way to identify biomedical research published in a preprint format...\" This is simply not true as Google Scholar indexes preprints. This was pointed out by Prachee Avasthi and in response the authors edited the text to include an incorrect method for finding preprints with Google Scholar. In a previous comment I pointed out how to correctly search for preprints with Google Scholar, and it appears the authors read the comment given they utilize the method at this page on their site: http://www.hsls.pitt.edu/gspreprints 3. In his comment the author states: \"We want to stress that the 'Sort by date' feature offered by Google Scholar (GS) is abysmal. It drastically drops the number of retrieved articles compared to the default search results.\" This feature of Google Scholar is indeed limited, as it restricts the results to articles which were published in the past year. However, if the goal is to find recent preprints then this limitation shouldn't be a problem and I don't know that I would classify the feature as \"abysmal\". 4. The article states: \"As new preprint servers are introduced, search.bioPreprint will incorporate them and continue to provide a simple solution for finding preprint articles.\" New preprint servers have been introduced, such as preprints.org and Wellcome Open Research, but search.biopreprint has not incorporated them. 5. Prachee Avasthi pointed out that the search.biopreprint search engine cannot find this F1000Research article about search.biopreprint. It only finds the bioRxiv version. In response the author stated: \"The Health Sciences Library System’s quality check team has investigated this issue and is working on a solution. We anticipate a quick fix of this problem.\" This problem has not been fixed."
}
]
},
{
"id": "14408",
"date": "06 Jul 2016",
"name": "Cynthia Wolberger",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article contains a thorough description of a new search tool, search.bioPreprint, that can be used to search multiple preprint archives using keywords. In light of the existence of multiple preprint servers that can be used to post preprints in the biological sciences, the development of a tool that enables full-text searching of all current preprint archives (arXiv, bioRxiv, F1000Research, PeerJ Preprints) is a welcome one. The search capabilities of search.bioPreprint and the bookmarklet app are well-described.\n\nLooking ahead, the authors will hopefully consider further improvements to the search site. Better documentation on the search site itself explaining how results are returned and ranked would be helpful. This reviewer quickly learned that the search returns approximate word match results, not just exact matches; this should be clarified on the web site. Moreover, approximate matches sometimes appear be ranked more highly than exact matches; the reason for this should be examined and remedied, if possible. It would also be helpful to have options to rank results in other ways, in particular, by date.",
"responses": [
{
"c_id": "2077",
"date": "20 Jul 2016",
"name": "Ansuman Chattopadhyay",
"role": "Author Response",
"response": "We sincerely thank the reviewer for considering the manuscript as “Approved” and for the constructive review. Reviewer Comment: “…the authors will hopefully consider further improvements to the search site. Better documentation on the search site itself explaining how results are returned and ranked would be helpful. “ Author’s Response: We appreciate the suggestions and will take measures to improve documentation on the search site. Reviewer Comment: “This reviewer quickly learned that the search returns approximate word match results, not just exact matches; this should be clarified on the web site. Moreover, approximate matches sometimes appear be ranked more highly than exact matches; the reason for this should be examined and remedied, if possible.” Author’s Response: We agree that better documentation is needed and will add a thorough description of the search process on the search site. One of the limitations of this search engine is it completely depends on the search prowess of the individual sources and we do not have any control over that. If the query term is placed within quotation marks, it forces the search engines to apply an exact word matching algorithm thus mitigating the issue of approximate word matching. Reviewer Comment: “It would also be helpful to have options to rank results in other ways, in particular, by date.” Author’s Response: We acknowledge the pressing need for an option to sort preprint articles by date as both reviewers mentioned it. We are actively working to bring this functionality into search.bioPreprint."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1396
|
https://f1000research.com/articles/5-1746/v1
|
19 Jul 16
|
{
"type": "Research Article",
"title": "Effect of gestational diabetes mellitus on maternal thyroid function and body mass index",
"authors": [
"Elrazi A. Ali",
"Hala Abdullahi",
"Duria A. Rayis",
"Ishag Adam",
"Mohamed F. Lutfi",
"Elrazi A. Ali",
"Hala Abdullahi",
"Duria A. Rayis",
"Mohamed F. Lutfi"
],
"abstract": "Background: The exact influences of thyroid functions on body mass index (BMI) are ill-defined in euthyroid pregnant women with gestational diabetes mellitus (GDM). Objectives: To investigate the effect of GDM on maternal thyroid functions and BMI. Methods: A case- control study was conducted in Saad Abualila Hospital, Khartoum, Sudan June to August 2015. Cases included women with GDM and healthy pregnant women as controls. Thyroid hormones [thyroid-stimulating hormone (TSH), free tri-iodothyronine (FT3), and free thyroxine (FT4)] and anti-thyroid peroxidase (anti-TPO) and anti-thyroglobulin (anti-TG) antibodies were measured. Results: BMI was significantly increased in GDM patients (26.3 (2.7) Kg/m2) compared with the control group (24.3(1.8) Kg/m2, P = 0.001). Levels of FT3 and FT4 were significantly decreased in GDM patients (0.632 (0.408 ─ 1.074) pg/ml; 0.672 (0.614 ─ 0.960) ng/dl) compared with the healthy pregnant women (0.820 (0.510─1.385) pg/ml, P = 0.021; 0.840 (0.767─1.200) ng/dl, P < 0.001). In contrast, anti-TPO and anti-TG were significantly higher in GDM patients (11.13 (7.969 ─13.090) IU/ml; 14.40 (10.91─20.69) IU/ml) compared with the control group (8.90 (6.375─10.48 IU/ml, P = 0.022; 10.50 (8.2─13.95) IU/ml, P = 0.010). BMI correlated negatively with FT3 (r = ─ 0. 375, P = 0.002) and FT4 (r = ─ 0. 316, P = 0.009) and positively with anti-TPO (r = 0.361, P = 0.002) and anti-TG (r = 0.393, P = 0.010). Conclusion: The present results add further evidence for decreased free thyroid hormones, increased anti-thyroid autoantibodies and higher BMI in patients with GDM compared to healthy pregnant women. BMI correlated directly with FT3 and FT4, but failed to demonstrate significant association with TSH.",
"keywords": [
"Body mass index",
"gestational diabetes mellitus",
"pregnancy",
"thyroid function"
],
"content": "Introduction\n\nAbnormal thyroid function and glucose tolerance have been both reported during pregnancy1–3. It was hypothesized that thyroid hormones gradually increase during the first trimester, but decrease gradually over the rest of pregnancy4–6. The steady rise of human chorionic gonadotropin (hCG) hormone during the first trimester was claimed to induce follicular thyroid cells to release of tri-iodothyronine (T3) and thyroxine (T4)7, which negatively feedback on thyroid-stimulating hormone (TSH)8. During the second and third trimesters, TSH increases while T3 and T4 decrease following hCG withdrawal. Lower levels of free T3 (FT3) and T4 (FT4) over the last two thirds of pregnancy can also be explained by high thyroid hormones transport proteins concentrations induced by placental estrogens9. Alternatively, higher levels of diabetogenic hormones, reduced physical activity, decreased energy expenditure, increased carbohydrate consumption, lack of adequate sleep and other stresses during gestation increase insulin requirements of pregnant women3. Increased insulin requirement enhances development of gestational diabetes mellitus (GDM) in susceptible pregnant women e.g. obese women10, and those with dysfunctional pancreatic β-cells11 or insulin resistance12.\n\nWeight gain is common among subjects with insulin resistance13 as well as those with hypothyroidism14. During pregnancy, the degree of insulin resistance seems to influence levels of thyroid hormones15 and pattern of change in maternal body mass index (BMI)16. In contrast to overt cases of thyroid disorders, the exact influences of T3 and T4 on BMI are ill-defined in euthyroid pregnant women17–20. According to Ashoor et al.,18, FT4 decreased while FT3 increased with higher BMI scores. Although paradoxical effects of FT3 and FT4 on maternal weight were also demonstrated in other reports17,19, some studies failed to reproduce these findings20. The euthyroid subjects studied by Milionis et al., did not show association between BMI and FT3 or FT4. However, the same study showed significant positive correlations between BMI and total T3 (TT3) as well as between BMI and total T4 (TT4) in females, but not males20. The present study aimed to investigate the effects of GDM on the maternal thyroid function and BMI. In addition, correlations between BMI, FT3 and FT4 were assessed to clarify how thyroid hormones affect maternal weight during pregnancy.\n\n\nMethods\n\nA case- control study was conducted in Saad Abualila Hospital, Khartoum, Sudan from June to August 2015. Pregnant women with singleton pregnancies who attended the hospital antenatal screening for diabetes mellitus were approached to participate in the study. After signing an informed consent, each pregnant woman was asked about her age, obstetric and medical profile. The weight and height were measured and BMI was calculated and expressed as weight (kg)/height (m)2. Women were excluded from the study if they were smokers, had history of hypertension and personal history of cardiovascular disease, had previous medical history of diabetes, were taking any medication (apart from iron supplementation), and had prior significant medical illnesses.\n\nCoustan and Carpenter12 criteria were adopted for the diagnosis of gestational diabetes, by which after a 100-g oral glucose load, two or more of the following plasma values were met or exceeded: fasting 95 mg/dl, 1 h 180 mg/dl, 2 h 155 mg/dl, and 3 h 140 mg/dl. Women with normal values were included as controls.\n\nVenous blood specimens (5 ml) were drawn from the median cubital vein and collected in vacutainer blood-collecting tubes. The tube specimens were allowed to clot and then were centrifuged for 10 min at 3,000×g to separate the serum which was stored at −20°C until analysis for thyroid hormones (TSH, free T3, and free T4) using immunoassay analyzer (AIA 360, Tosoh, Japan), following the manufacturer’s instructions. Specific anti-thyroperoxidase (anti-TPO) and anti-thyroglobulin (anti-TG) antibody profiles were analyzed using enzyme-linked immunosorbent assay (ELISA, Euroimmun, Lübeck, Germanykits).\n\nSPSS for Windows (version 16.0) was used for data analyses. Continue variables were checked for normality and their difference was compared between cases and controls using T-test and Mann-Whitney U, when the data were normally and not normally distributed, respectively. Spearman correlations were performed between the different variables. P < 0.05 was considered statistically significant.\n\nThe study received ethical clearance from the Research Board at the Department of Obstetrics and Gynaecology, Faculty of Medicine, University of Khartoum, Sudan.\n\n\nResults\n\nTable 1 shows the means (standard deviation, SD) of basic characteristics of the studied GDM patients and control group (34 women in each arm) including the age and gestational age. BMI was significantly higher in GDM patients (26.3 (2.7) Kg/m2) compared with the control group (24.3 (1.8) Kg/m2, P = 0.001).\n\nLevels of FT3 and FT4 were significantly decreased in GDM patients (0.632 (0.408–1.074) pg/ml; 0.672 (0.614–0.960) ng/dl respectively) compared with the healthy pregnant women (0.82 (0.510–1.385) pg/ml, P = 0.021; 0.840 (0.767–1.200) ng/dl, P < 0.001 respectively, Table 2). In contrast, anti-TPO and anti-TG were significantly higher in GDM patients (11.13 (7.969–13.090) IU/ml; 14.40 (10.91–20.69) IU/ml respectively) compared with the control group (8.90 (6.375–10.48 IU/ml, P = 0.022; 10.50 (8.2–13.95) IU/ml, P = 0.010 respectively), Table 2, Figure 1–Figure 5.\n\nBMI correlated negatively with FT3 (r = – 0.375, P = 0.002) and FT4 (r = – 0.316, P = 0.009) and positively with anti-TPO (r = 0.361, P = 0.002) and anti-TG (r = 0.393, P = 0.010).\n\nThere was no significant difference in TSH levels between GDM patients (2.037 (1.053–3.323) mIU//ml) and healthy pregnant women (2.401 (1.888–2.811) mIU//ml, P = 0.283) and no significant correlation with BMI (r = – 0.094, P = 0.446), Table 3.\n\n\nDiscussion\n\nIn accordance with the present results, previous reports demonstrated an associations between GDM and decreased thyroid hormones, increased anti-thyroid autoantibodies and higher BMI1,15,22–24. The associations between FT4, maternal weight, and GDM were recently investigated by Haddow and his group in more than 9000 singleton, euthyroid women in the FaSTER (First and Second Trimester Evaluation of Risk) trial2. An earlier report documented an inverse association between maternal weight and FT4 in the second trimester2. In a subsequent separate report on the same cohort, FT4 odds ratio for GDM was significant in the second (1.89), but not in the first (1.11) trimester1. Comparable findings were shown by Cleary-Goldman et al., when they demonstrated 1.7 odds ratio of hypothyroxinemia in GDM patients during the second trimester24. Oguz et al., confirmed decreased FT4 in 50 GDM patients compared with 60 non-GDM pregnant women; however, the mean of FT4 levels remained within the normal reference range in both groups15. Cases with isolated maternal hypothyroxinemia constituted 8% and 14% of GDM patients during the second and third trimesters respectively; however, similar cases were absent in the control group15. In another study, GDM patients showed lower FT4 compared to healthy pregnant women as well as those with type 1 diabetes mellitus. According to the same study, type 1 diabetic women had higher prevalence of anti-TPO compared with healthy pregnant women22.\n\nAccording to the present findings, BMI correlates negatively with FT3 and FT4, but positively with anti-TPO and anti-TG antibodies. In contrast, our results failed to demonstrate significant correlation between BMI and TSH. Increased odds of hypothyroxinemia and anti-TPO positivity among pregnant women with BMI ≥ 30 kg/m2 during the first 8 weeks of gestation was reported before25. Although several previous studies failed to establish an association between BMI and TSH after 8 weeks of gestation26–28, at least one study was able to do so when these two parameters were assessed during early pregnancy25. It was hypothesized that the peak of hCG hormone towards the end of the first trimester simulates simultaneous increase of thyroid hormones and reciprocal inhibition of TSH release7,8,18. Except for a temporal fall of TSH levels by the end of first trimester, both TSH and BMI steadily increase throughout pregnancy26,27. This explains why previous studies were able to prove significant positive correlation between BMI and TSH during the early 8 weeks of gestation25, but failed to reproduce same results during hCG surge26–28. However, failure of our results as well as other studies28 to document significant correlation between TSH and BMI during later stages of pregnancy is difficult to explain on the same basis and should motivate researchers in the field to investigate for possible explanation(s).\n\nAlthough the effects of thyroid hormones on body weight are easy predictable in cases with hypo- and hyperthyroidism, the influences of T3 and T4 on BMI are ill-defined in cases of euthyroidism26–28. The inverse relationship between FT4 and BMI demonstrated with the present results agreed with several previous reports17–19, but not others20. Likewise, the association between FT3 and BMI is a more contentious issue18,20. In a prospective cohort aimed to establish reference intervals of thyroid hormone concentrations among Finnish pregnant women, FT4 decreased while FT3 increased with higher BMI scores17. Same finding were reproduced by Ashoor et al., while assessing thyroid function before the start of the second trimester18. The paradoxical effects of FT3 and FT4 on maternal weight were further supported by Bassols et al., when they demonstrated significant direct association between FT3/FT4 ratio and BMI19. In contrast, a Greek study in euthyroid subjects failed to demonstrate the association between BMI and FT3 or FT4. The same study showed significant positive correlations between BMI and TT3 as well as between BMI and TT4 in females, but not males20. A possible explanation for different patterns of association between T3, T4 and BMI in previous reports is likely because of failure to adjust for confounders like caloric intake29,30. For example, conversion of T4 to T3 peripheral deiodinases is depressed in cases with caloric deprivation29 and enhanced with overfeeding30. This may explain the reciprocal effects of FT3 and FT4 and consequently the direct association between the FT3/FT4 ratio and BMI in cases with reduced caloric intake31. In well-fed states, peripheral deiodinase activity will not be augmented and consequently BMI is expected to correlate positively with thyroid hormones levels20. In conditions where FT3 and FT4 are below expected, increased TSH enhances leptin release and consequently BMI32. This may explain why FT3 and FT4 may negatively correlate BMI irrespective of caloric intake.\n\nOf note, deiodinase activity and insulin resistance were not assessed in the present study. Direct measures of deiodinase activity (e.g. hepatic deiodinase-1 mRNA) are difficult to evaluate because obtaining the required tissue samples is inconvenient. However, the T3/T4 ratio was proved to correlate well with deiodinase activity and can act as a surrogate for hepatic deiodinase-1 mRNA31. Evaluation of insulin resistance using parameters like Homeostasis Model Assessment (HOMA) in future studies will enable more clarification about the potential influence of insulin resistance on the relationship between BMI, FT3 and FT4. Another limitation of this study is that dietary composition and caloric intake were not evaluated among the studied women.\n\n\nConclusion\n\nThe present results add further evidence for decreased free thyroid hormones, increased anti-thyroid autoantibodies and higher BMI in patients with GDM compared to healthy pregnant women. BMI correlates positively with FT3 and FT4, negatively with anti-TPO and anti-TG antibodies, but failed to demonstrate significant association with TSH. Further studies that also evaluate deiodinase activity, caloric intake and indictors of insulin resistance are desirable for better understanding for the relationship between BMI, FT3 and FT4 in patients with GDM.\n\n\nConsent\n\nWritten informed consent to participate in the study and publish clinical details was obtained by the participants.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for effect of gestational diabetes mellitus on maternal thyroid function and body mass index, 10.5256/f1000research.9084.d12759933.",
"appendix": "Author contributions\n\n\n\nEAA and IA designed the study; HA and DAR carried out experimental protocols; IA, MFL analyzed and interpreted the data; MFL, IA wrote the first draft of the manuscript. All authors read and approved the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no supporting grants were involved in this work.\n\n\nReferences\n\nHaddow JE, Craig WY, Neveux LM, et al.: Free Thyroxine During Early Pregnancy and Risk for Gestational Diabetes. Crispi-Brillas F, ed. PLoS One. 2016; 11(2): e0149065. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaddow JE, Craig WY, Neveux LM, et al.: Implications of High Free Thyroxine (FT4) concentrations in euthyroid pregnancies: the FaSTER trial. J Clin Endocrinol Metab. 2014; 99(6): 2038–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDahlgren J: Pregnancy and insulin resistance. Metab Syndr Relat Disord. 2006; 4(2): 149–52. PubMed Abstract | Publisher Full Text\n\nMoon HW, Chung HJ, Park CM, et al.: Establishment of trimester-specific reference intervals for thyroid hormones in Korean pregnant women. Ann Lab Med. 2015; 35(2): 198–204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhalid AS, Marchocki Z, Hayes K, et al.: Establishing trimester-specific maternal thyroid function reference intervals. Ann Clin Biochem. 2014; 51(Pt 2): 277–83. PubMed Abstract | Publisher Full Text\n\nDhatt GS, Jayasundaram R, Wareth LA, et al.: Thyrotrophin and free thyroxine trimester-specific reference intervals in a mixed ethnic pregnant population in the United Arab Emirates. Clin Chim Acta. 2006; 370(1–2): 147–51. PubMed Abstract | Publisher Full Text\n\nGlinoer D: The regulation of thyroid function in pregnancy: pathways of endocrine adaptation from physiology to pathology. Endocr Rev. 1997; 18(3): 404–33. PubMed Abstract | Publisher Full Text\n\nGlinoer D: What happens to the normal thyroid during pregnancy? Thyroid. 1999; 9(7): 631–5. PubMed Abstract | Publisher Full Text\n\nCarayon P, Lefort G, Nisula B: Interaction of human chorionic gonadotropin and human luteinizing hormone with human thyroid membranes. Endocrinology. 1980; 106(6): 1907–16. PubMed Abstract | Publisher Full Text\n\nVillamor E, Cnattingius S: Interpregnancy weight change and risk of adverse pregnancy outcomes: a population-based study. Lancet. 2006; 368(9542): 1164–70. PubMed Abstract | Publisher Full Text\n\nBuchanan TA: Pancreatic B-cell defects in gestational diabetes: implications for the pathogenesis and prevention of type 2 diabetes. J Clin Endocrinol Metab. 2001; 86(3): 989–993. PubMed Abstract | Publisher Full Text\n\nKautzky-Willer A, Prager R, Waldhausl W, et al.: Pronounced insulin resistance and inadequate betacellsecretioncharacterize lean gestational diabetes during and after pregnancy. Diabetes Care. 1997; 20(11): 1717–1723. PubMed Abstract | Publisher Full Text\n\nMorrison JA, Glueck CJ, Horn PS, et al.: Pre-teen insulin resistance predicts weight gain, impaired fasting glucose, and type 2 diabetes at age 18–19 y: a 10-y prospective study of black and white girls. Am J Clin Nutr. 2008; 88(3): 778–88. PubMed Abstract\n\nLaurberg P, Knudsen N, Andersen S, et al.: Thyroid function and Obesity. Eur Thyroid J. 2012; 1(3): 159–167. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOguz A, Tuzun D, Sahin M, et al.: Frequency of isolated maternal hypothyroxinemia in women with gestational diabetes mellitus in a moderately iodine-deficient area. Gynecol Endocrinol. 2015; 31(10): 792–5. PubMed Abstract | Publisher Full Text\n\nCatalano PM: Obesity, insulin resistance, and pregnancy outcome. Reproduction. 2010; 140(3): 365–371. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMännistö T, Surcel HM, Ruokonen A, et al.: Early pregnancy reference intervals of thyroid hormone concentrations in a thyroid antibody-negative pregnant population. Thyroid. 2011; 21(3): 291–8. PubMed Abstract | Publisher Full Text\n\nAshoor G, Kametas NA, Akolekar R, et al.: Maternal thyroid function at 11–13 weeks of gestation. Fetal Diagn Ther. 2010; 27(3): 156–63. PubMed Abstract | Publisher Full Text\n\nBassols J, Prats-Puig A, Soriano-Rodriguez P, et al.: Lower free thyroxin associates with a less favorable metabolic phenotype in healthy pregnant women. J Clin Endocrinol Metab. 2011; 96(12): 3717–23. PubMed Abstract | Publisher Full Text\n\nMilionis A, Milionis C: Correlation between Body Mass Index and Thyroid Function in Euthyroid Individuals in Greece. ISRN Biomarker. 2013; 2013: 651494, 7. Publisher Full Text\n\nCoustan DR, Carpenter MW: The diagnosis of gestational diabetes. Diabetes Care. 1998; 21(Suppl 2): B5–8. PubMed Abstract\n\nVelkoska Nakova V, Krstevska B, Dimitrovski Ch, et al.: Prevalence of thyroid dysfunction and autoimmunity in pregnant women with gestational diabetes and diabetes type 1. Prilozi. 2010; 31(2): 51–9. PubMed Abstract\n\nAgarwal MM, Dhatt GS, Punnose J, et al.: Thyroid function abnormalities and antithyroid antibody prevalence in pregnant women at high risk for gestational diabetes mellitus. Gynecol Endocrinol. 2006; 22(5): 261–6. PubMed Abstract | Publisher Full Text\n\nCleary-Goldman J, Malone FD, Lambert-Messerlian G, et al.: Maternal thyroid hypofunction and pregnancy outcome. Obstet Gynecol. 2008; 112(1): 85–92. PubMed Abstract | Publisher Full Text\n\nHan C, Li C, Mao J, et al.: High Body Mass Index Is an Indicator of Maternal Hypothyroidism, Hypothyroxinemia, and Thyroid-Peroxidase Antibody Positivity during Early Pregnancy. Biomed Res Int. 2015; 2015: 351831,7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPop VJ, Biondi B, Wijnen HA, et al.: Maternal thyroid parameters, body mass index and subsequent weight gain during pregnancy in healthy euthyroid women. Clin Endocrinol (Oxf). 2013; 79(4): 577–583. PubMed Abstract | Publisher Full Text\n\nGowachirapant S, Melse-Boonstra A, Winichagoon P, et al.: Overweight increases risk of first trimester hypothyroxinaemia in iodine-deficient pregnant women. Matern Child Nutr. 2014; 10(1): 61–71. PubMed Abstract | Publisher Full Text\n\nHaddow JE, Craig WY, Palomaki GE, et al.: Impact of adjusting for the reciprocal relationship between maternal weight and free thyroxine during early pregnancy. Thyroid. 2013; 23(2): 225–230. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuda AK, Pittman CS, Shimizu T, et al.: The production and metabolism of 3,5,3'-triiodothyronine and 3,3',5-triiodothyronine in normal and fasting subjects. J Clin Endocrinol Metab. 1978; 47(6): 1311–9. PubMed Abstract | Publisher Full Text\n\nDanforth E Jr, Horton ES, O'Connell M, et al.: Dietary-induced alterations in thyroid hormone metabolism during overnutrition. J Clin Invest. 1979; 64(5): 1336–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgnihothri RV, Courville AB, Linderman JD, et al.: Moderate weight loss is sufficient to affect thyroid hormone homeostasis and inhibit its peripheral conversion. Thyroid. 2014; 24(1): 19–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZimmermann-Belsing T, Brabant G, Holst JJ, et al.: Circulating leptin and thyroid dysfunction. Eur J Endocrinol. 2003; 149(4): 257–71. PubMed Abstract | Publisher Full Text\n\nAdam I, Ali E, Abdullahi H, et al.: Dataset 1 in: Effect of gestational diabetes mellitus on the maternal thyroid function and body mass index. F1000Research. 2016. Data Source"
}
|
[
{
"id": "15785",
"date": "25 Aug 2016",
"name": "Rishi Ramtahal",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article studies the effect of gestational diabetes on maternal thyroid function and BMI. A case control study was done in Sudan and thyroid hormone levels and thyroid autoantibody testing was done.\nWith the increasing abnormalities in thyroid function such as subclinical disease seen in pregnancy, this study further consolidates existing literature with decreased thyroid hormone levels and increased thyroid autoantibodies in gestational diabetes.\n\nThis study also looked at the association between BMI and thyroid parameters. However, the conclusion states that BMI correlates positively with free thyroid hormones but the discussion says a negative correlation was present. Also the conclusion says BMI was negatively associated with thyroid autoantibodies but the discussion says the opposite. I assume this may be a typographical error in the conclusion section. Please address this issue.\n\nOtherwise this study, with its limitations adds interesting data to the preexisting literature on the subject.",
"responses": []
},
{
"id": "17047",
"date": "31 Oct 2016",
"name": "Tuija Männistö",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI read with interest the paper by Dr. Ali et al., describing a case-control study on thyroid function and autoimmunity during pregnancy among women with GDM. The study was small, but adequately powered to answer the study question - if women with GDM have poorer thyroid function or more thyroid antibodies than women without GDM. The statistical methods were adequately used in the study.\nMy concern with the methods was with the description of the control group, who are referred as healthy pregnant women. I would prefer that the controls would be described as women without GDM as the reader is not aware if these women have other pregnancy complications such as preeclampsia or fetal growth restriction. Another concern was the definition of euthyroidism - the Figure 1 clearly shows that some women had high TSH concentrations during pregnancy, some of which exceeded that of recommended euthyroid range specified by the American Thyroid Association, for instance. How did the authors define euthyroidism? The definition needs to be added to the Methods. Also, the authors should consider using a pregnancy-specific definition of euthyroidism, which may affect their results. At least they should do a sensitivity analysis after excluding women with TSH higher than pregnancy-specific reference ranges to see if their results hold. The observed results could be due to underlying hypothyroidism among women with GDM - this should also be discussed in more detail.\nAlso, it would be interesting to see a comparison between women with GDM with normal BMI and controls with normal BMI to see if GDM has an independent effect on thyroid function and autoimmunity irrespective of BMI. However, I wonder if the study population is big enough for this type of analysis.\nSpecific comments:\nAbstract In the Objectives please correct 'thyroid dysfunctions' to thyroid dysfunction. In the Methods please note that thyrotropin is a pituitary hormone, not a thyroid hormone. Please also add information on the statistical methods used in the Methods-section of the abstract. In the Results please add explanation as to what the figures are in the parentheses. Please also pay special note on the spacing between words, it seem some spaces are lacking.\nIntroduction The physiological changes in thyroid function during pregnancy are well described and accepted by the medical community. Therefore it was weird that the authors referred these as hypothesized and claimed (the first few sentences in the first paragraph). Could the authors reword these sentences?\nMethods Please add the conversion factor for plasma glucose levels to mmol/l. Please also add the laboratory method used to measure glucose, if available. If not available, please indicate if glucose was measured by a laboratory that participates in an external quality control program. In the Statistics section, please change 'Continue variables' to 'Continuous variables'.\nResults I presume that the gestational age in Table 1 is the timepoint when women entered the study? If so, please indicate this more clearly. Throughout the manuscript, please spell Kg as kg. Instead of decreased, I would use the word lower when describing differences between GDM cases and controls. Please round all results to two decimals as they are more meaningful clinically (also applies to Table 2).\nTable 1 Please add a space between the mean and SD. Please correct Kg/cm2 to kg/m2\nDiscussion Second line: please correct to 'an association' I think the study by Cleary-Goldman et al. studied the association between thyroid dysfunction and GDM, not the other way around (a subtle but important difference). I would love to see some discussion on the iodine status of the studied population.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1746
|
https://f1000research.com/articles/5-1736/v1
|
18 Jul 16
|
{
"type": "Software Tool Article",
"title": "cy3sabiork: A Cytoscape app for visualizing kinetic data from SABIO-RK",
"authors": [
"Matthias König"
],
"abstract": "Kinetic data of biochemical reactions are essential for the creation of kinetic models of biochemical networks. One of the main resources of such information is SABIO-RK, a curated database for kinetic data of biochemical reactions and their related information. Despite the importance for computational modelling there has been no simple solution to visualize the kinetic data from SABIO-RK. In this work, I present cy3sabiork, an app for querying and visualization of kinetic data from SABIO-RK in Cytoscape. The kinetic information is accessible via a combination of graph structure and annotations of nodes, with provided information consisting of: (I) reaction details, enzyme and organism; (II) kinetic law, formula, parameters; (III) experimental conditions; (IV) publication; (V) additional annotations. cy3sabiork creates an intuitive visualization of kinetic entries in form of a species-reaction-kinetics graph, which reflects the reaction-centered approach of SABIO-RK. Kinetic entries can be imported in SBML format from either the SABIO-RK web interface or via web service queries. The app allows for easy comparison of kinetic data, visual inspection of the elements involved in the kinetic record and simple access to the annotation information of the kinetic record. I applied cy3sabiork in the computational modelling of galactose metabolism in the human liver.",
"keywords": [
"Data display",
"Graphical user interfaces",
"Web service",
"SABIO-RK",
"kinetic parameters",
"SBML"
],
"content": "Introduction\n\nOne of the main challenges for the modeling of biochemical systems is the availability of reliable information on the individual reaction steps and their kinetics from the literature. This information includes kinetic parameters with their rate equations as well as detailed descriptions of how these were determined1.\n\nSABIO-RK (http://sabio.h-its.org/) is a manually curated database for kinetic data storing comprehensive information about biochemical reactions and their kinetic properties, with data manually extracted from the literature and directly submitted from lab experiments1,2. Available information comprises kinetic parameters with their corresponding rate equations, kinetic law and parameter types and experimental conditions under which the kinetic data were determined. In addition, information about the biochemical reactions and pathways including their reaction participants, cellular location and the catalyzing enzyme are recorded2.\n\nThe information in SABIO-RK is structured in datasets, so called database entries, which can be accessed either through the web-based user interface (http://sabiork.h-its.org/newSearch/index) or via web services (http://sabiork.h-its.org/sabioRestWebServices). Both interfaces support the export of the data in the Systems Biology Markup Language (SBML), a free and open interchange format for computer models of biological processes3.\n\nDatabase entries are annotated with controlled ontologies and vocabularies based on Minimum Information Required for the Annotation of Models (MIRIAM4), e.g. reaction participants (e.g. small chemical compounds and proteins), as well as kinetic rate laws, and parameters. The annotation information is encoded in the form of RDF-based MIRIAM annotations, and additional XML based SABIO-RK specific annotations, e.g. for experimental conditions. These annotations integrate the kinetic information with external resources like ChEBI (https://www.ebi.ac.uk/chebi/)5, UniProtKB (http://www.uniprot.org/)6, PubMed (http://www.ncbi.nlm.nih.gov/pubmed/), or KEGG (http://www.kegg.jp/)7.\n\nDespite the importance of kinetic information for computational modelling there has been no simple solution to visualize the database entries from SABIO-RK, and provide access to the network structure of the kinetic entries and the information encoded in the annotations.\n\nIn this work, we present cy3sabiork, an app for the visualization of kinetic data from SABIO-RK for Cytoscape, an open source software platform for network visualization8. cy3sabiork creates an intuitive visualization of kinetic entries in the form of a species-reaction-kinetics graph extended with kinetic information, which reflects the reaction-centered approach of SABIO-RK. Hereby, the kinetic information is accessible via a combination of graph structure and annotations of nodes, with provided information consisting of: (I) reaction details, enzyme and organism; (II) kinetic law, formula, parameters; (III) experimental conditions; (IV) publication; (V) additional annotations. cy3sabiork allows for easy comparison of kinetic data, visual inspection of the elements involved in the kinetic record and simple access to the annotation information of the kinetic record.\n\n\nMethods\n\ncy3sabiork was written in Java as an OSGi bundle for Cytoscape 3 using the app infrastructure. The bundle activator adds the cy3sabiork Action to the Cytoscape icon bar, which provides access to the JavaFX based cy3sabiork dialog. The cy3sabiork GUI is a combination of web based components handled in a WebView and classical GUI components. The combination of Swing and JavaFX is implemented based on a JFXPanel with JavaFX GUI updates in Platform.runLater, Swing GUI updates via SwingUtilities.invokeLater. The JavaFX approach allows for the integration of rich Web-based content using HTML/Javascript/CSS with Cytoscape. The GUI was created utilizing FXML based GUI definitions created in JavaFX Scene Builder, a JavaFX tool which allows to quickly design JavaFX application user interfaces by dragging UI components into a content view area. FXML code for the UI layout is created by the tool which was styled using CSS. The downside of the JavaFX-Swing hybrid approach are additional complexity in handling Events and EventListeners in different threads, some issues with Windows Management and Focus Handling, and full support of JavaFX in old felix OSGI containers.\n\nSABIO-RK entries are retrieved via the web services using the RESTful API. SBML from the web service calls or the web interface export is imported using a Cytoscape Task created by the LoadNetworkFileTaskFactory. CyNetworks and CyNetworkViews for the imported kinetic entries are created by a CyNetworkReader registered for SBML files provided by cy3sbml9. During the app development the SBML CyNetworkReader was extended to support the SABIO-RK specific annotations and data. RDF based annotations are read with JSBML10,11 and hyperlinks to the respective resources are created by parsing the resources.\n\nBy implementing cy3sabiork as a desktop app, in comparison to a solely web-based solution with Cytoscape.js, tight integration with cy3sbml was possible, thereby providing rich functionality for the kinetic entries in SBML, like for instance access to the annotation information or the raw SBML files of the kinetic entries.\n\nAn overview over the typical cy3sabiork workflow is depicted in Figure 1. The main steps of operation are searching entries in SABIO-RK, loading entries in cy3sabiork, and visual exploration of results:\n\nThe main steps of operation are: (1) Searching entries in SABIO-RK: Kinetic entries are retrieved from SABIO-RK via calls to the RESTful web services or via searching the web interface and exporting selected entries as SBML; (2) Loading entries in cy3sabiork: The exchange format between SABIO-RK and cy3sabiork is SBML; (3) Visual exploration of results. The kinetic graphs for the imported entries are generated providing simple access to kinetic information and annotations. An example query with resulting SABIO-RK information and subsequent visualization is shown in Figure 2.\n\nSearching kinetic entries. Kinetic entries can either be searched via the web services available from the cy3sabiork panel or directly in the SABIO-RK web interface available at http://sabiork.h-its.org/newSearch/index.\n\nThe web interface enables the search for reactions and their kinetics by either a free text search or an advanced search, which supports the creation of complex queries by specifying reactions by their participants (substrates, products, inhibitors, activators etc.) or identifiers (KEGG or SABIO-RK reaction identifiers and KEGG, SABIO-RK, ChEBI or PubChem compound identifiers), pathways, enzymes, UniProt identifiers, organisms (NCBI taxonomy12), tissues or cellular locations (BRENDA tissue ontology (BTO)13), kinetic parameters, environmental conditions or literature sources1,2. After finalizing the search the selected kinetic entries are exported as SBML.\n\nThe web interface supports the same query options, which can be added in the REST query GUI, but allows a direct import of the SBML without the requirement for search in the web interface and export and subsequent import of the SBML.\n\nExample queries are depicted in Figure 2 and Figure 4 with resulting SBML available as Supplementary File S1 and Supplementary File S2.\n\nThe kinetic entry 14792 for galactokinase (EC:2.7.1.6, UniProtKB:P51570) was retrieved via the web service query http://sabiork.h-its.org/sabioRestWebServices/kineticLaws/14792 (status 10-06-2016, SBML of query in Supplementary File S1). (A) Overview of kinetic information for SABIO-RK entry (http://sabiork.h-its.org/kineticLawEntry.jsp?viewData=true&kinlawid=14792) with color coding according to 1. (B) cy3sabiork information for entry 14792: (1) Resulting species-reaction-modifier graph. The galactokinase enzyme catalyzes the conversion of D-Galactose + ATP → α-D-Galactose 1-phosphate + ADP (see also Substrates, Products and Modifiers in A); (2) Kinetic graph with additional nodes for kinetic law, parameters and localization; (3) Selecting nodes in the graphs provides access to the annotation information and links to databases. In the example the kinetic law information is displayed in the Results Panel. (4) MIRIAM annotations with respective links to databases are available via the Results Panel; (5) Additional SABIO-RK annotations in XML for the experimental conditions are displayed in this section.\n\nFor human galactose metabolism 88 entries are available (status 28-06-2016, SBML of query in Supplementary File S2). For the selected Entry 14785 detailed information is provided on the right side.\n\nGraph of SABIO-RK kinetic information available for human galactose metabolism consisting of 88 entries (status 28-06-2016, SBML of query in Supplementary File S2). The kinetic graph consists of three clusters, separated based on the reported localization of the catalyzing enzyme. The lysosomal entries are non-canonical reactions in the galactose metabolism.\n\nLoading kinetic entries. The SBML exported from web interface searches is imported as in Cytoscape using cy3sbml9 (File → Import → Network → File). For queries to the web services the response SBML is imported automatically without the need for additional file operations.\n\nVisual exploration. The final step is the exploration of the kinetic entries in the species-reaction-modifier and the kinetic graph (e.g. in Figure 2 and Figure 4). The information from a wide range of resources and databases is integrated with the graph visualization of the kinetic records accessible as hyperlinks from the cy3sbml panel. Examples are the access to the source publication on PubMed from which the kinetic information was retrieved, the UniProtKB protein for which the kinetic information was measured, KEGG and ChEBI information for species involved in the reaction, or links back to the SABIO-RK database entry and reaction.\n\n\nUse cases\n\nWe applied cy3sabiork for kinetic parameter search and model construction of a kinetic model of galactose metabolism of the human liver (https://github.com/matthiaskoenig/multiscale-galactose) within the Virtual Liver Network (VLN) and Systems Medicine of the Liver (LiSyM) projects. A crucial step in building kinetic models of metabolism is the collection of kinetic information from the literature for the parameterization of the biochemical reactions. The search and visual exploration with cy3sabiork provides easy-access to a high-quality starting set of kinetic parameters and corpus of relevant publications for the processes of interest. Hereby, subsequent literature search and retrieval of referenced publications is simplified.\n\nA representative SABIO-RK query for galactokinase (EC:2.7.1.6, UniProtKB:P51570), the first step of hepatic galactose metabolization is depicted in Figure 2 retrieving the SABIO-RK kinetic record 14792. A more complex query used for parameter search is depicted in Figure 3 and Figure 4 retrieving all kinetic records available for human galactose metabolism.\n\nDuring model building publications with kinetic information not yet available in SABIO-RK were included in the database by the SABIO-RK curation service.\n\n\nConclusion\n\ncy3sabiork is a Cytoscape app for visualizing kinetic data from SABIO-RK providing the means for visual analysis of kinetic entries from SABIO-RK within their reaction context. Herby, the integration of kinetic parameters with computational models is supported. The availability of direct links to annotated resources from within the network context of the kinetic records provides important information for the knowledge integration with computational models.\n\n\nSoftware availability\n\nSoftware available from: http://apps.cytoscape.org/apps/cy3sabiork\n\nLatest source code: https://github.com/matthiaskoenig/cy3sabiork\n\nArchive source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.5742814\n\nLicense: GNU General Public License, version 3 (GPL-3.0): https://opensource.org/licenses/GPL-3.0",
"appendix": "Author contributions\n\n\n\nMK developed the app and wrote the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Federal Ministry of Education and Research (BMBF, Germany) within the research network Systems Medicine of the Liver (LiSyM, grant number 031L0054) and the Virtual Liver Network (VLN, grant number 0315741).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank the SABIO-RK team, the SBML community, and Cytoscape community for their support and help. A special thanks to the curation team of SABIO-RK including additional kinetic data in the database based on provided publications.\n\n\nSupplementary material\n\nSupplementary File S1: Kinetic entry 14792.\n\nSBML for query http://sabiork.h-its.org/sabioRestWebServices/kineticLaws/14792\n\nSupplementary File S2: Kinetic entries for human galactose metabolism.\n\nSBML for query: http://sabiork.h-its.org/sabioRestWebServices/searchKineticLaws/sbml?q=Pathway:%22galactose%20metabolism%22%20AND%20Organism:%22homo%20sapiens%22\n\n\nReferences\n\nWittig U, Rey M, Kania R, et al.: Challenges for an enzymatic reaction kinetics database. FEBS J. 2014; 281(2): 572–582. PubMed Abstract | Publisher Full Text\n\nWittig U, Kania R, Golebiewski M, et al.: SABIO-RK--database for biochemical reaction kinetics. Nucleic Acids Res. 2012; 40(Database issue): D790–D796. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHucka M, Finney A, Sauro HM, et al.: The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models. Bioinformatics. 2003; 19(4): 524–531. PubMed Abstract | Publisher Full Text\n\nLaibe C, Le Novère N: MIRIAM Resources: tools to generate and resolve robust cross-references in Systems Biology. BMC Syst Biol. 2007; 1: 58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Matos P, Alcántara R, Dekker A, et al.: Chemical Entities of Biological Interest: an update. Nucleic Acids Res. 2010; 38(Database issue): D249–D254. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUniProt Consortium: Ongoing and future developments at the Universal Protein Resource. Nucleic Acids Res. 2011; 39(Database issue): D214–D219. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanehisa M, Goto S, Furumichi M, et al.: KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 2010; 38(Database issue): D355–D360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKönig M, Dräger A, Holzhütter HG: CySBML: a Cytoscape plugin for SBML. Bioinformatics. 2012; 28(18): 2402–2403. PubMed Abstract | Publisher Full Text\n\nDräger A, Rodriguez N, Dumousseau M, et al.: JSBML: a flexible Java library for working with SBML. Bioinformatics. 2011; 27(15): 2167–2168. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez N, Thomas A, Watanabe L, et al.: JSBML 1.0: providing a smorgasbord of options to encode systems biology models. Bioinformatics. 2015; 31(20): 3383–3386. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSayers EW, Barrett T, Benson DA, et al.: Database resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2012; 40(Database issue): D13–D25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGremse M, Chang A, Schomburg I, et al.: The BRENDA Tissue Ontology (BTO): the first all-integrating ontology of all organisms for enzyme sources. Nucleic Acids Res. 2011; 39(Database issue): D507–D513. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKönig M: cy3sabiork: bugfix & DOI release. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "15057",
"date": "04 Aug 2016",
"name": "Barry Demchak",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper describes the cy3sabiork Cytoscape app, which enables users to query the SABIO-RK kinetics database from within Cytoscape and then formats the results as a Cytoscape network. It also allows a Cytoscape user to import an SBML file generated by the SABIO-RK web site via browser-based interactive use.\n\nThe paper is generally well structured and worded. My comments mainly address affordances that allow the paper to deliver more value via wording changes, formatting improvements, and the addition of critical additional information. These comments address small points that add up to real value to new readers and readers only semi-fluent with SABIO-RK.\nThe abstract is difficult to read because it lacks white space between paragraphs. Additionally, adding back in some missing articles (e.g., \"the\" and \"and\") would improve flow.\n\"cy3sabiork\" is a hard name to read ... readability can be improved with a phonetic pronunciation (... would this be si-three-sab-york ??)\nThe operation of cy3sabiork isn't clear from the paper. The web interface is described, but it would be helpful to have a tutorial describing step-by-step operation and giving a tour of the result of each step. This would be helpful both in this paper and on the App Store web site.\nThe Introduction proposes a visualization, but it's unclear what the visualization shows or why it was chosen. For someone trying to understand more about the value of this app, a short explanation would be valuable.\nThe Implementation section is detailed and sufficient. It would be very worthwhile adding text that claims this app as the first to render its UI using JavaFX, and discussing the pros and cons of this approach from both the developer and user perspective.\nIn the Operation section, it should be made clear that when the web interface generates SBML, it is written to a file that the user must then manually import into Cytoscape. Making this more apparent in the \"After finalizing the search ...\" sentence would be helpful, and avoiding the nominative case in the \"Loading kinetic entries\" section would make clear that the user is doing the work. Similar clarification in Figure 1 would be helpful, too.\nIn all figures, the captions are very useful but very long. According to F1000 style guide, these captions should be short. If the style is applied, the captions should be moved into inline text (with more discussion).\nIn figure 3, the separation between the form (on the right) and the screen shot (on the left) is unclear. It's hard to read this figure and understand it quickly.\nThe graph in figure 4 is too detailed for reproduction in a PDF. For users that want to study it more, it would be helpful if the supplementary material would contain the .cys file from which it was generated.",
"responses": []
},
{
"id": "16239",
"date": "12 Sep 2016",
"name": "Ali Salehzadeh-Yazdi",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe current study is a Cytoscape app which visualized kinetic data from SABIO-RK. The article and results are of interest, but overall the article will be largely improved if the novelty and aims of the app are clearer.\n\nThe abstract should be more focused and make shorter.\n\nThe introduction should be more concise and informative of previous works. A paragraph explaining that different approaches have been described the visualization of the kinetic data based on ……….. in …… platform. (and explain the highlights of each method).\n\nThe Figures are not in a high-quality resolution and figure 4 is not informative. The author just showed the main three clusters of human galactose metabolism. I suggest illustrating the method clearer (specially the operation).",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1736
|
https://f1000research.com/articles/5-1717/v1
|
15 Jul 16
|
{
"type": "Software Tool Article",
"title": "AutoAnnotate: A Cytoscape app for summarizing networks with semantic annotations",
"authors": [
"Mike Kucera",
"Ruth Isserlin",
"Arkady Arkhangorodsky",
"Gary D. Bader",
"Ruth Isserlin",
"Arkady Arkhangorodsky",
"Gary D. Bader"
],
"abstract": "Networks often contain regions of tightly connected nodes, or clusters, that highlight their shared relationships. An effective way to create a visual summary of a network is to identify clusters and annotate them with an enclosing shape and a summarizing label. Cytoscape provides the ability to annotate a network with shapes and labels, however these annotations must be created manually one at a time, which can be a laborious process. AutoAnnotate is a Cytoscape 3 App that automates the process of identifying clusters and visually annotating them. It greatly reduces the time and effort required to fully annotate clusters in a network, and provides freedom to experiment with different strategies for identifying and labelling clusters. Many customization options are available that enable the user to refine the generated annotations as required. Annotated clusters may be collapsed into single nodes using the Cytoscape groups feature, which helps simplify a network by making its overall structure more visible. AutoAnnotate is applicable to any type of network, including enrichment maps, protein-protein interactions, pathways, or social networks.",
"keywords": [
"network analysis",
"enrichment map",
"tag cloud",
"network clustering",
"complexity reduction",
"modular networks",
"annotations",
"cytoscape"
],
"content": "Introduction\n\nIdentifying clusters of nodes in a network, based on similarity of node attributes or connectivity between the nodes, is useful for defining groups of related nodes. For example, clusters in a protein-protein interaction network often represent molecular complexes, proteins that work together as a group to perform a specific function1, whereas clusters in a co-authorship network represent a group of authors that often collaborate and publish together.\n\nClusters can be used to create a visual summary of a network by drawing an enclosing shape around each cluster and adding a textual summary label next to each cluster (Figure 1). This technique is often effective in summarizing the results of a network analysis by highlighting main themes and categories within the network. AutoAnnotate (http://apps.cytoscape.org/apps/autoannotate) was originally created to aid pathway enrichment analysis using the Enrichment Map App2 as every enrichment map requires clustering and annotation, but AutoAnnotate is now being made available as a stand-alone app to benefit other types of analysis.\n\nA network is clustered, textual annotation associated with each cluster is automatically summarized as a single cluster label, and the results are visualized. Users can customize the view by selectively collapsing and expanding clusters.\n\nCytoscape enables users to add annotations on top of the network, including arrow, image, shape and text objects. However, fully annotating the clusters in a network can be cumbersome because the user must manually create and position each individual annotation. Furthermore, Cytoscape does not maintain any relationship between the network layout and the annotations, when the layout changes the user must manually reposition the annotations. Because of these limitations some users prefer to export their networks and add annotations using an external application, such as Adobe Illustrator.\n\nAutoAnnotate is a Cytoscape 3 App that identifies clusters and automatically draws shape and label annotations for each cluster. The enclosing shapes make it visually clear which nodes belong to each cluster when the clusters do not overlap. The generated labels provide a concise semantic summary of the data attached to the nodes in each cluster. AutoAnnotate maintains a relationship between the annotations and the nodes in a cluster, if the layout of the network changes then the annotations are automatically repositioned. AutoAnnotate maintains multiple sets of annotations for a single network, which allows the user to experiment with different clustering algorithms and label generation strategies. Additionally, AutoAnnotate allows clusters to be collapsed, which can simplify large networks by reducing potentially large sections of the network into single nodes.\n\nAutoAnnotate leverages two existing Cytoscape Apps to do most of its work: clusterMaker2 (http://apps.cytoscape.org/apps/clustermaker2) and WordCloud (http://apps.cytoscape.org/apps/wordcloud). The clusterMaker2 App provides several clustering algorithms and is directly called by AutoAnnotate to identify clusters of nodes in the network. The use of clusterMaker2 is optional, and the user may provide their own list of cluster identifiers. WordCloud is a Cytoscape App that creates a visual summary of selected attributes for a set of nodes by displaying a word tag cloud, where more frequent words are displayed using larger font size and adjacent words are grouped closer together. AutoAnnotate invokes WordCloud to generate a word tag cloud for the node data within each cluster, which is used to derive the text for the label annotations.\n\nAutoAnnotate can be installed from within Cytoscape via the App Manager or from the Cytoscape App Store.\n\n\nOperation\n\nThis section discusses a basic overview of the capabilities of AutoAnnotate. A more detailed user manual is available from http://baderlab.org/Software/AutoAnnotate.\n\nAutoAnnotate creates and manages a list of “Annotation Sets” for each network. An Annotation Set consists of a group of clusters and their associated annotations. Each network may have one active Annotation Set at time. The user may create as many Annotation Sets as they like, and can easily switch between them. The use-case for supporting multiple Annotation Sets is to allow the user to experiment with different clustering and summarization parameters, and then choose the most satisfactory resulting Annotation Set. Annotation Sets are saved to the Cytoscape session file.\n\nTo create an Annotation Set select AutoAnnotate > New Annotation Set… from the Apps menu. A dialog will pop up enabling the users to define the clusters in the network (Figure 2). Clusters can be defined in two ways: user defined or calculated automatically through AutoAnnotate. AutoAnnotate accepts any node attribute associated with the network to define clusters. Nodes with the same value for the attribute will be placed in the same cluster. This enables users to import cluster definitions from external programs or clustering algorithms. Alternately, clusters can be calculated within Cytoscape using the clusterMaker2 App3 which enables users to cluster a network using one of many clustering algorithms it provides. The clusterMaker2 algorithms are available as commands or from the Apps menu. Most of the algorithms can generate a cluster ID attribute in the node table, which can be consumed by AutoAnnotate. For convenience, AutoAnnotate also provides access to a subset of the clusterMaker2 algorithms directly from the Create Annotate Set dialog (Figure 2). This enables a user to quickly get started using AutoAnnotate without knowing the fine details of clusterMaker2. Currently, due to the simplified interface within AutoAnnotate, only clustering algorithms that take zero or one parameters and that run quickly are available directly from AutoAnnotate. After the clusters are computed, a label for each cluster is generated by the WordCloud app4.\n\nCluster options: ClusterMaker2 requires a clustering algorithm and edge weight attribute to be selected, otherwise a node attribute must be selected. Label options: User selects the node attribute to use for creating labels, the labelling algorithm to use, and the maximum number of words per label.\n\nOnce clusters and labels are calculated, AutoAnnotate automatically adds them as annotations to the network (e.g. ellipses around each cluster). Clusters are contained within bounding annotation shapes with its corresponding label as a text annotation directly above it. The look of the label and shape annotations may be modified using the Display Options panel (Figure 3). From this panel the user can adjust border width, shape type, opacity and visibility for shape annotations, and font size, font scaling and visibility for label annotations.\n\nChanging parameters here automatically updates the display.\n\nAfter an Annotation Set is created, it is shown in the main AutoAnnotate panel (Figure 4). This panel allows the user to choose which Annotation Set is currently active for each network. When an Annotation Set is active, a list of all its clusters and labels is displayed in the table. Each row represents one annotation (i.e. one cluster with its computed label) with the number of nodes it contains and whether it is collapsed or not.\n\nClusters and labels are shown. Clusters can be collapsed or expanded. An annotation can be customized via options available in a context sensitive (right click) menu.\n\nAlthough AutoAnnotate aims to automatically create the best labels and annotations, it also enables users to customize the resulting annotations. Clusters and labels can be manually adjusted. Multiple clusters can be combined. New clusters can be created from selected nodes directly in the network view. Labels can be re-generated automatically based on different label options. New Annotation Sets can be created by selecting a subset of clusters from an existing Annotation Set.\n\nTo further simplify complex networks AutoAnnotate has the ability to collapse some or all of the clusters automatically. The set of nodes belonging to a cluster are removed and replaced with a single group node representing the set. The group node is named using the computed label. When all clusters in a complex network are collapsed, it can substantially decrease the complexity of the network so that major themes or structure of the underlying network is more readily apparent.\n\n\nImplementation\n\nWordCloud registers a command service that calculates a word-cloud given a set of nodes and one or more node attributes. AutoAnnotate calls this command once for each cluster. The results of the command are placed in a list of WordInfo objects, which contain the font size of each word, the order that words appeared in the attributes (maintains word order), and the word adjacency group that contains the word. The word adjacency group is an identifier assigned to each word based on the number of times words occur in adjacent positions.\n\nWordCloud computes word frequencies which AutoAnnotate then uses to calculate the best combination of words to describe the given cluster, considering word frequency in the cluster compared to that in the entire network. Normally WordCloud returns all of the words in the cloud, however AutoAnnotate uses a maximum of 1–10 words to make a label, and therefore must make a decision of which words from the cloud to use and in which order. AutoAnnotate currently has two options for deciding this. The “Biggest Words” option sorts the words by font size, takes the N largest words, then sorts the result by word order (this preserves the original order that the words appeared in the selected attribute). The “Adjacent Words” option is a heuristic that attempts to balance word size with word adjacency information. First the words are sorted by font size, then a size bonus is added to every word that is in the same adjacency group as the N largest words. This causes words that are in the same group as the N largest words to be more likely to be chosen. The size bonus cannot cause a word to become bigger than the largest word in the group. Then the list is sorted again by size and the N largest words are selected. We have found through trial and error with many networks that a size bonus of eight results in labels that provide a good semantic description of the nodes in the cluster. Thus, we have made this the default label making option.\n\nThe Cytoscape group nodes feature is used to collapse clusters. When AutoAnnotate collapses a cluster it first creates a group node that contains all the nodes in the cluster and then the group node is collapsed. The shape and label annotations are no longer drawn for the collapsed cluster. When the cluster is collapsed Cytoscape will create \"meta-edges\" between the group node and any other nodes it is connected to. The collapsed group nodes and the meta-edges provide a summary of the network. When a cluster is expanded the group node is first expanded and is then deleted. Collapsed groups must be expanded before switching to another Annotation Set because the nodes contained in each group may belong to different clusters in the other Annotation Set.\n\nAutoAnnotate is a Cytoscape bundle App based on the Cytoscape 3-supported OSGi (Open Services Gateway Initiative) (https://www.osgi.org/developer/specifications/) module framework. AutoAnnotate version 1.1 depends on Java 8, the Cytoscape 3.3 API and WordCloud 3.1. The Cytoscape App Manager or the Cytoscape App Store installs WordCloud automatically when installing AutoAnnotate. ClusterMaker2 is not installed automatically because it is optional. To make installation of clusterMaker2 easier for the user the Create Annotation Set dialog attempts to detect the presence of clusterMaker2, and if not available the user is presented with a web link to the App Store page for clusterMaker2 from which they can install it.\n\nCytoscape makes its extensive API available to apps via a number of OSGi service interfaces. These services must be acquired in a bundle activator lifecycle method that is called when the App is initialized. Traditionally the service references are passed down the chain of constructors that build the App’s object graph. This technique is known as Dependency Injection, since a class’ dependencies are “injected” when the class is instantiated rather than the class having the responsibility of looking up the dependencies it needs. Dependency Injection is a well known software design pattern that improves modularity and facilitates unit testing. However, in Cytoscape this Dependency Injection is traditionally done manually, by hand-coding constructor parameter lists, which is verbose and error prone. AutoAnnotate uses the Google Guice Dependency Injection framework (https://github.com/google/guice/wiki/GettingStarted) along with the Peaberry OSGi adapter (https://github.com/google/guice/wiki/OSGi). Guice is configured to acquire Cytoscape services through Peaberry, and those services are then made available to all objects instantiated through Guice, removing the need to hand code constructor method parameter lists. The result is cleaner code that is easier to maintain and unit test.\n\nThe internal architecture of AutoAnnotate is divided into three main modules and some smaller supporting modules. The three main modules are: 1) Data Model, which is controlled by the Model Manager and contains the model for Annotation Sets, clusters, display options and labels. The Data Model is saved to the Cytoscape session file. 2) The UI, which consists of the main panel (Figure 4), the display options panel (Figure 3), the create annotation set dialog (Figure 2), menu items, and various other information and warning dialogs. 3) The Annotation Renderer, which draws the label and shape annotations on the network canvas.\n\nThe Model Manager listens to Cytoscape events, such as events for nodes being selected, and modifies the Data Model to stay in sync with Cytoscape. The Model Manager emits Model Events whenever the Data Model changes, caused by reacting to Cytoscape events or caused by user interaction. Client modules that listen to model events do not use the traditional observer pattern5, which is ubiquitous in object oriented user interface programming, instead the EventBus class from the Google Guava library is used. An EventBus allows event subscribers to be fully decoupled from event sources since each source does not need to maintain a listener list. The Model Manager emits the Model Events over the EventBus and the other main modules (UI and Renderer) subscribe to the EventBus and declare which event types they want to receive. Subscribing to the EventBus is very simple since it is injected by Guice. When the user interacts with a control in one of the UI panels a mutator method is called on the data model, this in turn fires a model event, which causes the rest of the UI to be updated. The UI and network view are kept in sync with the model without introducing any direct dependencies between the UI panels, the annotation renderer or the data model. Figure 5 depicts the event flow.\n\n1) Cytoscape events are fired using OSGi services (e.g. SetCurrentNetworkViewEvent, NetworkViewAboutToBeDestroyedEvent). 2) The AutoAnnotate Model Manager reacts to Cytoscape events and updates the Data Model. 3) The ModelManager fires Model Events over the Guava EventBus (e.g. AnnotationSetAddedEvent, ClusterAddedEvent). 4) The UI and Annotation Renderer modules each respond to a subset of the model events. There is no direct dependency between the UI and the Annotation Renderer. 5) The UI directly updates the Model, which causes Model Events to fire, which updates the UI and Annotations.\n\nIntegration with the WordCloud and clusterMaker2 Apps is made possible via the Cytoscape command-executor API (http://chianti.ucsd.edu/cytoscape-3.4.0/API/org/cytoscape/command/package-summary.html). ClusterMaker2 registers a command service for each of the clustering algorithms it provides. AutoAnnotate calls these commands using the CommandExecutorTaskFactory service.\n\n\nResults and discussion\n\nWe present four use cases for AutoAnnotate. The first, the most common use case, is an enrichment map. Clusters in an enrichment map represent similar pathways and processes. AutoAnnotate offers a way to easily summarize the main themes present in the network and is used routinely by enrichment map users. The second use case further demonstrates AutoAnnotate summarization capabilities by demonstrating how an annotated network can be collapsed to just its major themes. The third use case demonstrates how AutoAnnotate can be used for a different type of network, a co-authorship network. Finally, the fourth use case demonstrates another network type, a protein-protein interaction (PPI) network annotated with Gene Ontology (GO) terms6, to demonstrate how multiple annotation sets for an individual network can be used to summarize the same network to get different results.\n\n\nUse case 1 - Enrichment maps\n\nAn enrichment map is a graphical representation of a pathway enrichment analysis where nodes represent pathways and edges the crosstalk (or shared genes) between connecting pathways. To illustrate this, we downloaded gene expression from the TCGA Ovarian serous cystadenocarcinoma RNASeq V2 cohort on 2015-05-22 from cBioPortal for Cancer Genomics (http://www.cbioportal.org/data_sets.jsp). Enrichment analysis was conducted on a subset of this data as described in the enrichment map protocol7. As an example, the resulting enrichment map was annotated using AutoAnnotate (Figure 6A).\n\n(A) Annotated enrichment map (B) Collapsed enrichment map.\n\nEnrichment results from an ovarian cancer gene expression analysis are depicted as an enrichment map (q-value<0.0001, enrichment analysis performed as per EM protocol7). The resulting enrichment map was annotated using parameters - WordCloud normalization of 0.5, MCL clustering, and pathway description was used for label calculation (Figure 6A). WordCloud calculates the word frequencies of the attribute selected. For an enrichment map we chose the pathway description attribute to calculate labels as it contains a concise description of each node. The normalization factor controls how the word frequencies are weighted compared to the words in the rest of network. With a normalization factor of zero the significance of each word is calculated solely on how many occurrences it has in the given cluster. This may cause very frequent words within the network such as “pathway” or “regulation” to be prominent in annotations as they are found often in pathway descriptions. By increasing the normalization factor, we increase the weight calculated from the ratio of a word frequency in the cluster to its frequency in the entire network to diminish the presence of these recurrent words in the cluster labels.\n\n\nUse case 2 - Collapsed AutoAnnotated enrichment map\n\nIn an enrichment map visualization (and networks of other types) our gaze often gravitates towards the largest most densely connected region or cluster. In an enrichment map such clusters do not necessarily represent the most important themes of the analysis but instead biological pathways that are well studied and documented in public databases. AutoAnnotate allows users to normalize this bias by collapsing the network. Both large and small clusters become a single node labelled with the computed annotation and nodes not part of any cluster retain their original pathway or node name. In this format large clusters are no more important than single nodes but it is easy to quickly see the major themes of an analysis even within large complicated networks. The enrichment map was further summarized by collapsing all clusters to individual group nodes that represent the cluster and are named according to the annotation given. To Collapse the network simply select “Collapse All” from the AutoAnnotate Input panel menu (Figure 6B).\n\n\nUse case 3 - Co-authorship network\n\nAutoAnnotate is not limited to enrichment maps. Any standard network can be annotated based on node attributes in the network. For example, co-authorship networks where nodes are authors and edges connect two authors that have published together can also be annotated by AutoAnnotate. Using all the publications that had Cytoscape in its title/abstract/keyword as extracted from Scopus8 using the SocialNetworkApp9 we constructed a co-authorship network. The large network was filtered to contain only authors that had more than 20 citations. The resulting network (Figure 7) was annotated using the ‘publications’ attribute that contains the titles of every article an author published (in this case any article that contained the word ‘cytoscape’ in its title/abstract/keywords for the given author). The WordCloud normalization factor was set to 0.3 (slightly smaller than the previous enrichment map use case). Given that the set of articles dealt exclusively with Cytoscape they shared many similar words and themes. A high normalization factor would have eliminated most of these similar words like ‘network’, ‘analysis’ and ‘biological’. We wanted to highlight the different areas that use Cytoscape as well as the prominent purposes. Experimenting with the WordCloud normalization factor allowed us to find the ideal balance. The annotated network was collapsed to show the varied themes of the articles relating to Cytoscape.\n\nThe most central node was annotated with an additional image representing the top words in the set of publications as generated by wordle.com.\n\n\nUse case 4 - PPI network\n\nAutoAnnotate enables the user to create multiple annotations based on different parameters and easily toggle between the different annotations. Figure 8 shows the annotation of a PPI network generated from co-purified complexes10. A subset of complexes was annotated using different attributes including protein names, GO biological process, molecular function or cellular component. Depending on the attribute used to compute annotations the summarization gives a slightly different view of the data.\n\nFour different annotations for the same base PPI network using parameters - WordCloud normalization of 0, previously created cluster designations, and A. protein names, B. GO biological process, C. GO molecular function, or D. GO cellular component was used for label calculation.\n\n\nConclusion\n\nAutoAnnotate aids in network analysis and interpretation by providing a concise visual summary of clusters in a network. Clusters can be visualized either as shape and text annotations overlaid on top of the network, or by collapsing the clusters into single nodes.\n\nUsers can get started quickly with AutoAnnotate by accepting the defaults when creating an Annotation Set. More advanced users can experiment with different strategies for generating clusters and labels. Users are free to specify clusters themselves manually or using other clustering tools.\n\nAutoAnnotate fits into a larger workflow to help analyze a network or help present the results of an analysis. For example AutoAnnotate is an important part of the enrichment map protocol.\n\n\nSoftware availability\n\nHomepage: http://baderlab.org/Software/AutoAnnotate\n\nSoftware available from: http://apps.cytoscape.org/apps/autoannotate\n\nLatest source code: https://github.com/BaderLab/AutoAnnotateApp\n\nArchived source code as at the time of publication:\n\nhttps://zenodo.org/record/57021#.V3vOgpMrJE411\n\nLicense: Lesser GNU Public License 2.1: https://www.gnu.org/licenses/old-licenses/lgpl-2.1.html",
"appendix": "Author contributions\n\n\n\nMK and RI wrote the manuscript. RI provided experimental data to validate the use of the app. AA developed the initial prototype version of the app and provided input on the user interface design. MK developed the final version of the app. GB supervised app development and provided input on the manuscript.\n\n\nCompeting interests\n\n\n\nThere are no competing interests.\n\n\nGrant information\n\nThis work was supported by NRNB (U.S. National Institutes of Health, grant number P41 GM103504) and R01 GM070743.\n\n\nAcknowledgements\n\n\n\n\nReferences\n\nBader GD, Hogue CW: An automated method for finding molecular complexes in large protein interaction networks. BMC bioinformatics. 2003; 4(1): 2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerico D, Isserlin R, Stueker O, et al.: Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One. 2010; 5(11): e13984. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorris JH, Apeltsin L, Newman AM, et al.: clusterMaker: a multi-algorithm clustering plugin for Cytoscape. BMC bioinformatics. 2011; 12(1): 436. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOesper L, Merico D, Isserlin R, et al.: WordCloud: a Cytoscape plugin to create a visual semantic summary of networks. Source Code Biol Med. 2011; 6(1): 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVlissides J, Helm R, Johnson R, et al.: Design patterns: Elements of reusable object-oriented software. Reading: Addison-Wesley. 1995; 49(120): 11. Reference Source\n\nAshburner M, Ball CA, Blake JA, et al.: Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000; 25(1): 25–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEnrichmentMap protocol. (in preparation).\n\nBurnham JF: Scopus database: a review. Biomed Digit Libr. 2006; 3(1): 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKofia V, Isserlin R, Buchan AM, et al.: Social Network: a Cytoscape app for visualizing co-authorship networks [version 3; referees: 1 approved, 2 approved with reservations]. F1000Res. 2015; 4: 481. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHavugimana PC, Hart GT, Nepusz T, et al.: A census of human soluble protein complexes. Cell. 2012; 150(5): 1068–1081. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKucera M: AutoAnnotateApp: AutoAnnotate 1.1.0. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "16447",
"date": "04 Oct 2016",
"name": "Pablo Porras Millán",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a new Cytoscape app, AutoAnnotate, that integrates topological cluster search with automatic annotation in order to help visualize underlying structures in complex networks. The tool comes to complement previous work by the same authors in other Cytoscape apps such as EnrichmentMap or WordCloud. This app provides a much-needed tool for semi-automatic annotation and visualization of complex network data.\n\nI found the app particularly useful when used to visually annotate and summarize term enrichment analysis results, which was its original design motivation. Used in combination with EnrichmentMap it provides users with enhanced visualization options on popular tools, such as BiNGO, g:Profiler or DAVID.\n\nThe manuscript is clearly and concisely written, providing a nice introduction on the app functionality, its motivation and how to start using it. The tutorial referenced in the text is also quite useful, I have followed it to test the app with different use cases and I could get satisfactory displays without spending much time tweaking my results. The examples shown were also relevant and helped showing the app capabilities, although I would have welcomed an example using third party tool data, such as sorting out the network visualization output of BiNGO.\n\nTo summarize, this is a valuable tool that I hope to see much used in network annotation visualization in the future.",
"responses": []
},
{
"id": "16577",
"date": "24 Oct 2016",
"name": "Harm Nijveen",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAutoAnnotate is a Cytoscape app that can make complex networks easier to interpret by clustering nodes and labelling the clusters with relevant terms. Clusters can be collapsed into single nodes, greatly simplifying the network. The app is easy to use and allows configuration of quite of number of settings. The clusters are labelled based on the most frequent words in a (configurable) node column, using functionality from the WordCloud app.\nThe proper labelling of the clusters can require some toying with the parameters and manual curation, both in AutoAnnotate and WordCloud. For instance, using the AlzheimerEM.cys network from the WordCloud tutorial a cluster was labelled with the terms “interphase cycle”, where “cell cycle” would be more appropriate. The word “cell” was omitted due to the default “Normalize” setting of the WordCloud app that prevented very frequent words to be present in a label. Lowering this setting from 0.5 to 0.3 changed the label to “cell cycle”. This is described in the quite elaborate AutoAnnotated User Guide.\nI have also tested the app with the SeedNet network (http://netvis.ico2s.org/dev/seednet/#/static/Data) which is a co-expression network that has 8,621 nodes with 502,173 interactions. The nodes were pre-clustered with MCODE into 136 clusters. It took AutoAnnotate a few minutes to annotate the clusters (on a MBP 2012), which is acceptable, and then about 10 seconds to generate a summary network. The produced cluster labels were based on the annotation of the clustered genes. The labels were not very informative because there were hardly any common words in the annotations. This was obvious from the words in the labels, and from the output of the WordCloud app for a cluster. For a future version of the app it might be an idea to scale the font of the label by the frequency of the chosen words (now the font can be scaled by the cluster size), to indicate how descriptive a label is for a certain cluster.\nThe manuscript is well written and to the point.\nIn the following two lines the plural “they” refers to the singular “user”: “The USER may create as many Annotation Sets as THEY like, and can easily switch between them.” “if not available the USER is presented with a web link to the App Store page for clusterMaker2 from which THEY can install it.”\nAutoAnnotate is a useful addition to the collection of Cytoscape apps, and will undoubtedly help many users to create an informative summary of a complex network.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1717
|
https://f1000research.com/articles/5-1714/v1
|
15 Jul 16
|
{
"type": "Software Tool Article",
"title": "webANIMO: Improving the accessibility of ANIMO",
"authors": [
"Willem Siers",
"Michiel Bakker",
"Bob Rubbens",
"Ruben Haasjes",
"Jacco Brandt",
"Stefano Schivo",
"Willem Siers",
"Michiel Bakker",
"Bob Rubbens",
"Ruben Haasjes",
"Jacco Brandt"
],
"abstract": "ANIMO is a Cytoscape 3 app to model biological signalling pathways. Useful analyses can be performed and displayed to the user in an effective way. However, all this power comes at a cost: the additional software requirements for ANIMO have been hindering its widespread adoption. Our goal has been to provide beginner to intermediate ANIMO users with a simpler and more effective platform to perform their research: webANIMO. The minimalistic interface provides everything the regular ANIMO user needs for the most common tasks. Adding the fact that it is a web interface removes any software requirements from the equation. This article describes how webANIMO works: its client/server architecture, how Cytoscape and ANIMO compatibility was maintained, the visualization techniques implemented and other general design decisions.",
"keywords": [
"network dynamics",
"systems biology",
"visualization",
"regulatory networks",
"ANIMO",
"web interface",
"timed automata"
],
"content": "Introduction\n\nThe desktop version of ANIMO1,2 is currently implemented as an app for the popular biological network visualization program Cytoscape3. Users are able to create networks modelling biological processes in the Cytoscape interface. A node (also known as a reactant) in the network usually describes a protein or a gene, and has an activity level that is the ratio between the active and inactive form of the reactant. The edges connecting reactants are called interactions, and describe how the reactants influence each other: following biological custom, → means activation and ⊣ stands for inhibition. These networks are analysed in ANIMO using the UPPAAL4 tool in order to calculate the changes in activity levels over time. These changes are represented as a graph and reflected on the network using node colors to represent different activity levels.\n\nANIMO was developed in collaboration with biologists, and is used in biological studies5,6. While ANIMO is made to be user-friendly and make the biologist feel “at home”, the initial impression may not be entirely positive due to the requirements for its installation. The need to install Java, Cytoscape and UPPAAL before being able to install ANIMO and finally see whether it suits their needs could be considered as a hurdle for ANIMO’s target user group. For this reason, we decided to develop a web-based version of ANIMO that would make it much easier to access the main features of the tool. The idea of webANIMO was thus implemented in a design project at the University of Twente.\n\n\nMethods\n\nOne of the most important aspects of the architecture of webANIMO (see Figure 1) is the task division between client and server. Since web browsers have limited memory and processing power, the tasks of the client are kept to a minimum. The main task for the browser is visualization, while all the calculations are kept server-side and are delegated to a headless version of ANIMO. Cytoscape JS 2.6.17 is used to handle the visualization of networks. However, in order to give the web version the same functionality as the desktop version of ANIMO, many additions and modifications were necessary. webANIMO gives the user most of the features that the desktop version also provides, for example, the ability to add, edit and delete nodes or edges in the network and visualize the changes of activity level over time, using a graph and changing the colours of nodes. Major features of ANIMO that are not included in webANIMO are parameter fitting, model checking and a history of previous simulations.\n\nSince the interface of the desktop version of ANIMO is relatively complex, the decision has been made to make the web interface different, but effort has been spent to maintain all the features of the desktop version. The interface of the web version in general still resembles the desktop version. The upper section of the page is used for a taskbar, which contains submenus. The left of the page is used for the legend and the analyse options. The rest of the page is used by the network view and the result view. The width of these two views can be adjusted to the users need. Figure 2 and Figure 3 show the desktop and the web versions, respectively, for visual comparison.\n\nThe editor has been built using Cytoscape-JS7, which provides the basic functionality for loading a network and displaying the nodes and the edges. The functionality to add, edit and delete nodes and edges has been added, using jQuery 1.129 in order to make it easier to program this functionality. Cytoscape-JS provides events for tapping on nodes and edges: these events hide and show the Edit window. Changes made are shown instantaneously. jQuery change() events have been bound to all the input fields that can be edited in order to provide instant feedback to the user. For example, when the user changes the activity level of a node, the colour will change simultaneously. The webANIMO editor also provides undo- and redo-functionality to let the user manage changes to the network. Furthermore, the webANIMO editor has support for hot keys: users are able to use common hot keys for undo, redo, escape, enter, zoom-in and zoom-out.\n\nWhen first starting with webANIMO, the only thing presented to the user is an empty network. In order to better guide the user, the editor has a toolbar with all features: the user can press the buttons in order to add nodes and edges, or to zoom and fit the network to the view. Moreover, the buttons serve as a quick guide. The qTip 2.2.1 library10 was used to create clear tool tips: when the user hovers over the tool bar buttons, quick tips about using the editor are shown.\n\nThe result view was built using the D3 line-chart plugin11. After a network has been analysed, the webANIMO front-end receives from the back-end the analysed data in JSON format, which is then presented to the user as a graph. Similar to the desktop version of ANIMO, the graph has a slider on top which allows the user to see the activity levels of nodes at different time points: while the user drags the slider along the graph, nodes in the network are colored depending on their activity level at the corresponding time points. A legend is shown at the bottom of the graph: by clicking on the relevant node, users are able to hide and show plot data. The graph also has zoom features, allowing the user to take a closer look at specific points of the graph in more detail.\n\nOne of the earliest requirements of webANIMO was that it must be compatible with desktop ANIMO so that users can switch back and forth between the online and desktop environment without too much effort and technical know-how.\n\nTo make this possible, the functionality of importing and exporting of Cytoscape session files was implemented. Import and export functionality gives the user the option to choose the appropriate working environment. If the user prefers the improved UI and does not mind the reduced feature set of webANIMO the user can avoid installation issues and choose webANIMO. Once the user needs more processing power or more advanced features, the user can choose to export their progress to the native Cytoscape format and continue working in desktop ANIMO. Furthermore, once the user decides (s)he is finished with the advanced feature set of desktop ANIMO and wants to move back to the minimal interface of webANIMO, (s)he is free to do so.\n\nFor the sake of usability, the import feature can handle both Cytoscape 2 and Cytoscape 3 session files. For the same reason it was decided to export to the Cytoscape 2 format, because this format is also supported by Cytoscape 3 and it was a simpler export target. The concrete functionality of importing and exporting was written in Javascript and is executed completely client side, which makes it fast even given bigger networks since the networks do not have to be uploaded. This also reduces the workload of the server. The downside to this design choice is that if one of the file format specifications changes, the import feature might need to be updated.\n\nWhen the user presses the ‘Analyse’ button, a request containing the current model is sent to the server, which simulates that model and sends a response back containing the results of the simulation. The UI is disabled during the analysis and the waiting time is limited to one minute, after which the simulation is forced to terminate and no result is sent back. The use of a simulation time-out also reduces server load. The choice to perform the analysis on the server was made so the user is not required to install the tools that are needed to perform the simulation. As an added benefit this may improve the simulation time for the user if the device sending the request has slower hardware than the server (which is usually the case if the user accesses webANIMO through a tablet or smartphone).\n\nThe data is sent to the server with an HTTP POST request, containing request data formatted as JSON, as shown in Figure 4 and Figure 5. This JSON document has two required fields, “model”, which describes the biological network, and “minutesToSimulate’, which specifies how many minutes should be simulated. The PHP program that handles the POST request sends the JSON document to the ANIMO back-end Java program. This program performs the simulation and formats the simulation results as JSON, which is printed to standard output. The PHP program reads the results and sends them back to the client.\n\nThe code that performs the simulation is an extension of the desktop version of ANIMO. A JSON input is converted to an ANIMO network object, where all fields in the “data” JSON object are converted to properties in the ANIMO network. The back-end of webANIMO is an addition to the original ANIMO source code, so any changes to the simulation logic will be supported by webANIMO as well.\n\nANIMO transforms the input model to a timed automata model, which is simulated with UPPAAL 4.14. The output of the simulation is a textual description of the changes in node activity levels over time. This result is converted to a JSON document similar to the one shown in Figure 6, and sent back to the client.\n\nUsers of webANIMO may desire other ways of analysing the results of a simulation. To facilitate this, a feature to export the data to the CSV (comma-separated values) format was added. This allows the users to import the results into, for example, a spreadsheet editor and perform additional calculations or visualise the data. In the exported CSV data, there is a column with time values in minutes, and a column for every plotted node with the activity levels at the corresponding minute. It is possible that one node has an activity value at some time point, where another node does not. In this case the missing value is omitted.\n\nIn order to confirm that the model of a biological process produces the same results as in real life, biologists can import experimental data in the CSV format. This data is added to the plot window together with the analysis data. For consistency, the format of the imported CSV data must be the same as the exported CSV data.\n\nWhen a user is working with webANIMO, (s)he would not want to lose their work when they accidentally close their browser. To prevent this loss, webANIMO saves the current state of the network on the user’s device whenever a change is made. The last saved state is loaded again when one revisits the webANIMO application. This functionality was implemented using the new HTML5 web storage API12. To account for web browsers that do not yet fully support this API, webANIMO uses Mozilla’s backwards compatibility code13, which relies on cookies to store data locally.\n\nwebANIMO plots can also be exported as images. The plot export feature works on the basis that the plot is inherently a Scalable Vector Graphics (SVG) element. This means that the plot is easily modifiable in the browser and it can be converted to a wide range of image formats. These two reasons allow for a robust implementation of exporting the SVG plot to a downloadable image. For an improved user experience the user can choose to directly download an image, or show the exported image in the browser first. The plot export feature is implemented as follows. First the SVG plot element is cloned, to avoid changing the actual plot in the interface. Then, a legend is added, several superfluous elements are removed, and the element as a whole is resized to accommodate a slightly bigger font. Then the SVG is rendered to an HTML5 canvas, which subsequently is exported to the PNG (portable network graphics image) format. This is then, depending on the choice of the user, either presented in a new tab or available for download directly.\n\nAlthough webANIMO is designed to be as intuitive as possible, it might be possible that a potential user does not immediately see all modelling possibilities by just “playing” with their own networks. To provide the user with a more complete overview of everything one can do with webANIMO, the application has packed two example models. The examples can be found as entries of menu-item ‘Examples’: they are called ‘Model base’ and ‘Chrondrocyte’ and were presented in 14 and 5 respectively.\n\nWhen the user changes the current network not only the new network is saved, but also the change that the user performed is saved. During a session (which is ended when the page is reloaded) these changes are accumulated and called a revision. In the “show revisions” view (Figure 8) these revisions are listed and can be clicked in order to view the changes that happened in each of them. The changes are visualized instead of just displayed as text, to allow visualisation of all the new changes in a quick glance. Modifications to edges and nodes are indicated with light blue, creations with green, deletions with red and the unchanged sections of the graph are grey. This makes collaborative work on bigger networks simpler to manage.\n\nNode “Erk” is selected.\n\nwebANIMO allows to share networks with other people. The networks made in webANIMO will by default be saved to the browser’s local storage, so that when reloading or revisiting webANIMO the same network is loaded again. However, when a user wants to share the current network with another user, the network (including its revisions and changes) is stored on the web-server. When the network is first sent to the server for saving, the server creates a random (unused) filename and stores the network in a file with this name. This filename is a random 128-bit value in hexedecimal form, so that guessing filenames is not feasible. The user is then redirected to ./?network=N, where N is the filename of the file that stores the network. This filename is also the network identifier and is unique. The user can then share this link with other users: upon opening the link the associated network is loaded. The changes to the network are regularly saved to the file and can be seen as revisions (Figure 8).\n\nTo prevent multiple users editing the same network simultaneously and overwriting each other’s changes, a lock file is created on the server for each network. In this lock file the ip-address of the user editing the file is stored. When opening a link for a network a comparison is performed on the last time the lock was used and the current time. If the file is recently used by another ip-address, the network will be loaded, but an error is displayed indicating that changes will not be saved to the server.\n\nTo run webANIMO, the user needs any internet connected device with an (up-to-date) modern browser. Browsers such as Chrome 51, Firefox 47, Safari 9.1, Edge 25 and Internet Explorer 9+ are capable of running webANIMO, although it has been most extensively tested on Chrome and Firefox.\n\nInstalling webANIMO on a server requires PHP version 5.2 or higher, Java version 8 or higher and an UPPAAL 4.1 installation.\n\n\nUse cases\n\nWe will now briefly describe the process of creating a simple network with two reactants and one interaction connecting them.\n\nFirst create a reactant by right clicking on an empty location in the network view or selecting the “Add reactant” button in the toolbar. The property panel will change to “Add reactant” mode. Give the reactant a name and click the “Save” button to add the reactant to the network view.\n\nThe reactant should now be visible somewhere in the network view. If not, click the “Zoom to fit” shortcut in the toolbar to find it. Now add another reactant to the network view in the same way as before. Now there will be two reactants visible in the network view.\n\nIn order to add an interaction, click the “Add new interaction” button in the toolbar, which will cause the property panel to change to the “Add interaction” mode. As the source reactant, select the reactant that was first created. As the target reactant, select the reactant that was created after that. Make sure the “K” slider in the propery panel is on the “Slow” position, then click save.\n\nA network with two reactants and one interaction should now be visible in the network view. In order to get an interesting analysis result, one more thing needs to be done. Click on the source reactant of the interaction.\n\nThe property panel will change to the “Edit reactant” mode. Change the “Initial activity level” slider to the highest level, then click close.\n\nThe network is now ready to be analysed. Click on the “Analyse” button in the property panel, after which a “Loading, please wait” dialog will appear. The analysis of the network is now being done on the webANIMO server. Once the analysis is complete the dialog will disappear and the analysis result will be displayed in the plot view. The red vertical bar in the plot can be clicked and dragged along the plot to display the activation levels of the reactants at that point in the simulation.\n\nMore instructions on how to perform basic functions in webANIMO are described in the webANIMO manual, which can be found under the “Help” menu.\n\n\nSummary\n\nThis paper presents an online tool to model and analyze biological networks, based on an existing Cytoscape app. Most of the original features of the app (ANIMO) are now available simply accessing a website, which eliminates the initial installation phase that has been described as hideous and difficult by our target audience of biomedical students. The process of integrating this Cytoscape plugin into a web application has been described in this paper, and makes use of the CytoscapeJS library. Using this library and the described approach, other Cytoscape plugins could also be ported if there is a need to.\n\n\nSoftware availability\n\n1. Software available from:\n\nhttp://fmt.cs.utwente.nl/tools/webANIMO\n\nThe webANIMO Manual can be found at\n\nhttp://fmt.cs.utwente.nl/tools/webANIMO/manual.pdf\n\n2. Latest source code:\n\nhttps://bitbucket.org/willemsiers/webanimo_frontend\n\n3. Archived source code as at time of publication:\n\nhttp://dx.doi.org/10.5281/zenodo.5720619\n\n4. License:\n\nMIT (https://opensource.org/licenses/MIT)",
"appendix": "Author contributions\n\n\n\nWS, MB, BR, RH and JB worked on the development of webANIMO and the writing of the associated documents. SS supervised the development of webANIMO and provided feedback on the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSchivo S, Scholma J, Wanders B, et al.: Modeling biological pathway dynamics with timed automata. IEEE J Biomed Health Inform. 2014; 18(3): 832–839. PubMed Abstract | Publisher Full Text\n\nSchivo S, Scholma J, Wanders B: Cytoscape 3 animo app.2016. Reference Source\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLarsen KG, Pettersson P, Yi W: Uppaal in a nutshell. Int J Softw Tools Technol Transf. 1997; 1(1): 134–152. Publisher Full Text\n\nScholma J, Schivo S, Urquidi Camacho RA, et al.: Biological networks 101: Computational modeling for molecular biologists. Gene. 2014; 533(1): 379–384. PubMed Abstract | Publisher Full Text\n\nScholma J, Schivo S, Kerkhofs J, et al.: ECHO: the executable chondrocyte. In Tissue Engineering & Regenerative Medicine International Society. European Chapter Meeting, Genova, Italy. Malden, Wiley, 2014; 8. : 54–54. Reference Source\n\nFranz M, Lopes CT, Huck G, et al.: Cytoscape.js: a graph theory library for visualisation and analysis. Bioinformatics. 2016; 32(2): 309–311. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchivo S, Wanders B: Analysis of Networks with Interactive MOdeling.2016. Reference Source\n\nThe jquery foundation. 2016. Reference Source\n\nThompson C: qtip2.2016. Reference Source\n\nBostock M: Data-driven documents (d3).2016. Reference Source\n\nInc. World Wide Web Consortium: Editor Ian Hickson, Google. Web storage specification, 2015.2016. Reference Source\n\nMozilla Developer Network and individual contributors: Local storage compatibility code.2016. Reference Source\n\nSchivo S, Scholma J, Karperien HBJ, et al.: Setting parameters for biological models with ANIMO. In É André and G. Frehse, editors, Proceedings 1st International Workshop on Synthesis of Continuous Parameters, Grenoble, France. volume 145 of Electronic Proceedings in Theoretical Computer Science. Open Publishing Association; 2014; 35–47. Publisher Full Text\n\nAanstoot D, Beets F, Bolhaar G, et al.: Bringing ANIMO to the future of biology.2015.\n\nSchivo S: Design project form-animo.2015.\n\nGoogle Inc: AngularJS.2016. Reference Source\n\nHolt M. 2016. Reference Source\n\nSiers W, Bakker M, Rubbens B, et al.: webANIMO front end source code. Zenodo. 2016. Data Source"
}
|
[
{
"id": "15146",
"date": "01 Aug 2016",
"name": "Denis Thieffry",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI) Summary\nThe authors reports the development of a web interface, called webANIMO, for their former Cytoscape app ANIMO, which allows the dynamical modelling and analysis biological signalling networks. This client/server application removes the burden technical installation burden, thereby providing naive users direct and easy access to the tool and some of its analysis features.\nII) Suggestions on the manuscript\n1) We would suggest to reorganize the figures to improve their general organisation and emphasize more the features of webANIMO with respect to ANIMO:\nMerge Figures 1 and 4, keeping two distinct parts.\n\nSend Figure 2 to the supplementary material, as this manuscript is about webANIMO, and I don't think that it is very useful to display the interface of ANIMO in the core of this manuscript, and anyway its contents is largely redundant with that of Figure 3.\n\nSend Figure 5 to the supplementary material or to the tutorial, as its contains is likely too technical for the naive user of webANIMO, and probably not very informative for the expert.\n\nThe Figure 7 appears of limited interest and partly redundant with the right part of Figure 3. Perhaps, it would be more interesting to use this figure to illustrate another case study.\n\n2) As UPPAAL plays a central role, it would be good to include a brief introduction to this tool and more specifically to the functionalities (and underlying principles) used in webANIMO.\n3) It is not clear to us what are the restrictions regarding the kinetic terms that can be encoded with webANIMO vs ANIMO. The set of “scenarios” considered seems to be pretty limited. Their naming refer to biochemical reactions whatever the type of molecules defined as sources and targets. Furthermore, their number appears limited and their nature very simple... Would it be possible for example to consider Hill functions (widely used for transcriptional regulations)? The consideration of sufficiently non-linear terms is a prerequisite to the reproduction of sophisticate dynamics, such as oscillatory behaviour. In any case, these restrictions should be clearly stated in the manuscript, perhaps b y including a table with examples of reactions and interactions (in isolation or in combinations) along with recommended scenarios and descriptions of their limitations.\n4) It would be useful to include a few more case studies, e.g. implementations of simple gene networks such as the toggle switch (Gardner et al., 2000) and the repressillator (Elowitz et al., 2000).\nII) Testing the tool Web interface\nThe web interface could be opened successfully and was fast to load. No particular error was observed while randomly clicking on the interface. No problems were observed while importing/exporting networks or plots. The exported session could successfully be locally imported in Cytoscape.\nProposed improvements/new functionalities\n5) Default name should be proposed for any new reactant (e.g. Reactant1, Reactant2, etc.), while having the text selected for direct edition by the user. This would allow fast creation of Reactants in order to focus more on the structure of the network.\n6) To delete a node, apart from the right-click option, one has to use the button \"Remove\" which is close to the Description text area, thus difficult to find. It is not intuitive that this button deletes reactants. A more intuitive way to implement this functionality would be a \"Delete\" icon in the menu (usually a garbage bin) removing selected elements.\nUnwanted behaviors\n7) From an empty network, create two reactants and start adding a reaction between them. While the \"Add interaction\" panel is active it is still possible to delete the reactants through right click menu: the thin blue reaction line is still displayed on the editor with no possibility to remove it excepted from reloading the page\n8) From a network, use Zoom to fit and then Zoom out a few times. The displayed network will become partially hidden from the view, obliging the user to re-center it. A more intuitive behavior would be to zoom using the centre of the editor as reference, instead of the top left corner.\n9) Apparently, the \"Network\" -> \"New\" functionality actually clears the current network even if the network was shared. This might create some confusion in the case of shared networks, as one user could clear accidentally a shared network. The previous version would not be available through the Revision history. It might be more intuitive to (i) add a \"Clear\" item in the \"Network\" menu, (ii) create a new URL when the user clicks \"Network\" -> \"New\".\n10) When clicking on \"Network\" -> \"Get shareable link\" a text message warns the user that his network is going to be saved remotely, but there is no \"Cancel\" option in case this was not intended.\nManual\n11) Section 2 of the Manual presents basic features, but no visual help or guidance through screenshots, while this would help new users find their way in the interface and see if what they are doing is correct or not.\n12) webANIMO comes with two example networks \"Model base\" and \"Chrondrocyte\". It would be helpful for the new user to have a short description of these networks in the Manual, as well as expected results after direct simulation through with the \"Analyse\" button. Proposed exercises on these networks could help the new user experiment and learn specific features in webANIMO.\n13) I would suggest to split the current Manuel into two complementary documents: a Manual and a tutorial: - The manual aiming to provide a comprehensive description of the tool and its functionalities. - The tutorial aiming to help the user to develop his first models and analyses, starting from simple applications up to more complex ones.",
"responses": []
},
{
"id": "15047",
"date": "26 Aug 2016",
"name": "Gary D Bader",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors introduce a new Cytoscape.js-based web app that enables users to simulate a given biological network based on a range of parameters, taking advantage of most of the features of ANIMO (originally developed as a Cytoscape app) without the requirement of installing Cytoscape on the desktop. In general, the new webANIMO system seems useful and well developed, but we have a number of suggestions to improve the manuscript.\n\nGeneral comments:\nThe paper mixes descriptions of the user interface with implementation, which is confusing for the reader. It would be better to start with the analysis and use case sections, showing a flowchart of the overall workflow as figure 1, followed by a section describing the user interface (UI) and then the implementation details in a separate section. In particular, the meaning of a ‘model’ should be explained. Also, the ‘analysis’ should be described. What is involved? What is the input and what is the output?\n\nThe authors claim that webANIMO was developed because users find the ANIMO app difficult to install. They even go so far as to describe the process as “hideous and difficult”. This claim should be better supported e.g. by providing results of user uptake or satisfaction between ANIMO and webANIMO. ANIMO installation does not seem terribly difficult.\n\nAbstract: “Useful analyses can be performed and displayed to the user in an effective way.” Be more specific in what useful analyses webANIMO is for.\n\nThe webANIMO URL should be presented more prominently (e.g. at the top), as it currently only appears at the end of the text.\n\nIntroduction section: The introduction should start with an explanation about what ANIMO and webANIMO are for.\n\nActivation and inhibition symbols are defined but don’t seem to be used again in the text – is this just explaining the network visualization representation? In this case, it should be in a figure legend.\n\nThe UPPAAL system should be briefly described.\n\nImplementation Section: What is the purpose of the front end and what is the purpose of the back end? What information is shown/processed at each end? An overall figure explaining the information flow would be useful.\n\nWhat does it mean to run ‘headless’ ANIMO? Is this implemented as a standalone java process, or is it running inside of the Cytoscape OSGi container? This is an important distinction because the ANIMO app depends on Cytoscape. Was that dependency removed?\n\nUser Interface section It is mentioned that the decision was made to make the web interface “different”. Different in what way? The very next sentence says the user interface still resembles the desktop version. These sentences seem to contradict each other.\n\nResult view section: A more detailed description of the meaning of the result view and what it shows would be useful.\n\nImport and Export section: The ability to work in webanimo and desktop animo seamlessly is a great feature. It could be pointed out that this will help Cytoscape users share information with collaborators who may not have installed Cytoscape.\n\nIt is mentioned that import/export is done client side to avoid server workload. This contradicts the statement in the implementation section that “Since web browsers have limited memory and processing power, the tasks of the client are kept to a minimum”.\n\nIt is mentioned that users can switch to desktop ANIMO when they need more processing power. How powerful are the webANIMO servers and when should users move to the desktop version? In general, the authors should clarify when to use the desktop vs. web-based versions and what are the capabilities and features supported by each.\n\nWhat sort of network size (number of nodes and edges) does this limit support? At what point do users need to move to the cytoscape desktop version? What are the limits of the server? (All of these details should be added to the implementation section)\n\nAnalysis: Clarify description of time throughout the paper – server processing time or simulated concentration over time? Why was the limit of one minute chosen? Is this one minute of processing time on the server or does it also include data transfer?\n\n“The UI is disabled during the analysis and the waiting time is limited to one minute, after which the simulation is forced to terminate and no result is sent back.” Why does the platform not return any partial results?\n\n“The choice to perform the analysis on the server was made so the user is not required to install the tools that are needed to perform the simulation. As an added benefit this may improve the simulation time for the user if the device sending the request has slower hardware than the server (which is usually the case if the user accesses webANIMO through a tablet or smartphone).” This sounds useful, however, the web platform became non-responsive on mobile devices while the simulation is running during testing.\n\nA useful feature not available in webANIMO is the possibility to change the \"Reaction kinetics\" (available in the desktop version). Is this feature expensive in term of computational costs? Looking at the code, this feature has been implemented (embedded in the HTML code) and disabled. Why did the authors decide to not enable this feature?\n\nPlot Export section: Clarify meaning of graph (network vs. plot/chart) throughout the manuscript.\n\nRevision History section In the implementation section it is mentioned that one of the features of ANIMO that was not implemented in webANIMO is “history of previous simulations”. This section describes a feature called “revision history”. How are these features different?\n\nOnline shared networks section: What are the benefits of saving and sharing a network vs. just saving the cytoscape session and sending the session? Is there anything in the saved and shared network that is not in the session?\n\nWith the lock files, if you share the network with someone who opens it but then leaves it open in the browser (potentially for hours or days) does that lock you out of your own network?\n\nIt is nice how the user is given the choice to store their data locally or on the server.\n\nOperation section: \"Installing webANIMO on a server requires PHP version 5.2 or higher, Java version 8 or higher and an UPPAAL 4.1 installation.\" – clarify why one would want to install a local version vs. using the easy to use web version or the Cytoscape app.\n\nuse cases section: This is more like a simple tutorial than a list of use cases. Suggest either describing use cases only (e.g. gene knock out), develop the tutorial better (e.g. adding figures), or replace with a link to an online tutorial.\n\nMinor points: \"propery\" should be property\n\nFigure 1: this picture could be improved by moving the \"Cytoscape JS\" block inside the \"webANIMO\" block and the \"webANIMO\" block inside the \"web browser\" block. Also, remove \"JS\" from the webANIMO block name, since webANIMO is platform, not a library. Or, if the JS is meant to indicate that the block is implemented in javascript (and not a module name), then the languages for the other blocks should be specified for consistency. In the latter case, “Cytoscape JS” should be “cytoscape.js”\n\nFigure 4: \"POST\" could be removed from the I/O system diagram – the main point is request/response.\n\nWe noticed a Javascript error associated with the tooltip library (when interacting with the line chart)",
"responses": []
},
{
"id": "15453",
"date": "30 Aug 2016",
"name": "Aurélien Naldi",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWebANIMO is a web interface to the ANIMO modelling tool, enabling to get a grasp of the tool while avoiding a complex installation process (the desktop version is a cytoscape plugin which further requires the UPPAAL software). The client uses cytoscape-JS and provides a model editor, relying on JSON requests to the server to perform simulations. Some features are only available in the desktop version, import and export of cytoscape session files enable the user move between the two versions. The paper explains the motivation behind this web version, and the architecture of the client and server parts. The tool itself and the information available in the paper are sound, but the paper could include more context and be streamlined.\nSome background is missing, especially on ANIMO itself and UPPAAL: a complete description of these tools would be out of scope, but a brief introduction would improve the paper.\nA paragraph outside of the methods to introduce the user interface and the feature set (edit models, see their revision history, share them online and exchange with the desktop version. Perform analysis on the server, visualize and export the result) would help to understand the scope.\nFigures 1 and 4 can probably be merged, providing a single view of the client-server architecture. In figure 1, what is the difference between \"ANIMO\" and \"headless ANIMO\" ? Also, does the server version rely on cytoscape (I understand that it doesn't but it could be more clear) ?\nFigures 5 and 6 could be better suited for supplementary material or code documentation: stating that client-server communication uses JSON is enough for many technical readers, and these figures will probably not help the non-technical ones.\nThe authors provide a demo server for webANIMO, with some restrictions on running time to save resources. What would you encourage users who need more power to do: use the desktop version or install their own server? Installing their own server sounds like a good option for groups with both bioinformatics resources and several naive users.\nTrying out the web interface:\nThe tool loaded properly and seems to be working well despite the following minor issues.\nIf I start to add an interaction with a right-click on the graph, and select a node without validating the new interaction, the temporary link is still visible but can't be added to the model.\nI encountered a few CSS issues, which could be tied to my browser (firefox 48 on fedora):\nthe \"initial activity slider\" is there but not visible (clicking on the gray area under the label does work). The \"k\" slider for interactions had the same problem. the text in the comboboxes was too large and only readable after opening the menu.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1714
|
https://f1000research.com/articles/5-1699/v1
|
14 Jul 16
|
{
"type": "Software Tool Article",
"title": "SCODE: A Cytoscape app for supervised complex detection in protein-protein interaction graphs",
"authors": [
"Sarah Mohamed",
"Nick Janus",
"Yanjun Qi",
"Sarah Mohamed",
"Nick Janus"
],
"abstract": "Protein complexes are groups of interacting proteins unified by a common biological function. Identifying complexes amid a network of thousands of interacting proteins poses a difficult computational challenge. Traditional approaches to this problem rely on clique-like topography in order to identify complexes. Supervised learning is an alternative approach that leverages real-valued data in order to extract the features of protein complexes and identify candidates that do not conform to traditional, dense clique structures. SCODE (Supervised Complex Detection), an application for the Cytoscape App Store, implements a supervised learning algorithm for the detection of protein complexes in protein-protein interaction networks.",
"keywords": [
"supervised learning",
"complex",
"Bayesian network",
"protein-protein interaction",
"search"
],
"content": "Introduction\n\nProtein-protein interaction (PPI) networks provide information about the biochemical relationships in a cell’s molecular machinery. Each node in the network represents a protein, with edges connecting proteins that physically interact with one another. Complexes are clusters of interacting proteins in the network which are together responsible for some biological function. The discovery of complexes in a PPI network has implications in the study of the molecular basis of diseases, drug targets, and biological pathways1.\n\nExisting tools for complex detection in PPI networks rely on assumptions about a common topography observed among most protein complexes. These tools often assume that protein complexes can be identified based on the density of interactions (edges) among the set of its nodes. As a result, the problem is often reduced to clique detection2. While many complexes often take the form of edge-dense protein clusters, this is not the only observed topology. Real-valued data indicate that a variety of other features, including but not limited to edge density, are typical of complexes among PPI networks3.\n\nSCODE, an application for supervised complex detection, seeks to make more informed predictions about potential clusters in PPI networks. It applies supervised learning, a strategy for making predictions based on training data that has already been labeled (as a complex or a non-complex, for example). In order to predict novel complexes, SCODE uses a naive Bayesian classifier to determine the likelihood that potential protein clusters are true complexes.\n\nSCODE allows the user to define a set of features that determine whether a protein cluster forms a complex or not. The probability distribution of this set of features, dependent on whether or not a cluster is a complex, is encoded using naive Bayesian networks. Based on information in the training data, this probability distribution determines which predicted protein clusters conform to the observed features of known complexes. Not only does this identify more qualified candidate complexes, it also provides the user with more autonomy to describe what characteristics are important.\n\nOverall, SCODE performs four major functions: (1) train a Bayesian network, (2) search a PPI graph for candidate complexes, (3) score candidate complexes, and (4) optionally, evaluate the quality of the returned complexes. Figure 1 illustrates the pipeline of these functions in SCODE.\n\n\nMethods\n\nThe search algorithm uses a form of iterative simulated annealing in which complexes are expanded to include neighboring proteins at each cycle. The user may specify a set of \"seed\" proteins from which to begin the search, or allow the program to randomly select a set of starting nodes.\n\nBeginning from each of the seed proteins, the program iteratively performs the following steps:\n\n1. Maintain a record of the neighbor that produces the highest score when added to the complex. For each neighbor of the seed:\n\n(a) Calculate the complex score if the neighboring protein is added\n\n(b) If the complex score exceeds the record, then update the record’s value\n\n2. If the record exceeds the current score of the complex, then add the associated protein to the complex\n\n3. If the record does not exceed the current score of the complex, then calculate the probability of adding the protein to the complex as a function of the score and the temperature. SCODE then adds the protein to the complex with probability as follows: P(exp(Record – Score(complex))/Temperature)\n\nThree variations on this algorithm are available: iterative simulated annealing (ISA), greedy iterative simulated annealing (Greedy ISA), and sorted-neighbor iterative simulated annealing (Sorted-Neighbor ISA). In ISA, at each iteration a node is randomly selected from the neighbors of the candidate complex and scored in order to determine if it should be kept in the complex. Greedy ISA scores all neighbors of the candidate complex and retains only the best one (or none, if no neighbors improve the score). Sorted-Neighbor ISA sorts the neighbors of the complex by degree and evaluates the top N neighbors (N is a parameter set by the user).\n\nSCODE offers two scoring options to the user. The first does not perform learning, while the second option does (by training a Bayesian model).\n\nThe first scoring option performs a simple calculation based on the mean weight among edges in the proposed protein cluster. It does not require any information regarding known complexes and does not perform learning. This option may be used as a reference for comparing the complexes produced by the second scoring method.\n\nThe second scoring option, relying on supervised learning, calculates scores using a likelihood equation that is generated based on a trained Bayesian network. It reflects the conditional probability distribution that each feature is demonstrated in a complex versus the distribution for a non-complex. The sample Bayesian network from Figure 2 would use the following calculation to determine the likelihood that a candidate protein cluster, x, represents a true complex, given the prior probability that x is a complex (P(x1)) or a non-complex (P(x0)):\n\nThe root indicates whether the candidate cluster is a complex or not.\n\nHere P(f1|x1) denotes the conditional probability that feature f1 is observed given that x is a complex (x1). P(f1|x0) indicates the conditional probability that feature f1 is observed given that x is not a complex (x0). This equation may be extended to any naive Bayesian network, with f1…fn representing the features of the model:\n\nBayesian networks (BNs) are directed acyclic graphs (DAGs) that describe the conditional dependency relationships among a set of nodes representing random variables. In SCODE, Bayesian networks are used to describe the joint probability distribution of the features describing complexes, as well as the joint probability distribution of the features describing non-complexes in the PPI graph. After training the networks to calculate each of the conditional probabilities in Equation 1, these networks are used to score complexes that are discovered during the search phase.\n\nA feature describes a particular property of a candidate complex, such as size or density. Each node in the trained Bayesian model represents a discrete value, or subset of values, of a feature. For example, a binary feature is represented in the trained model using two nodes, with one node for each of its two possible values. A feature with continuous values is discretized by subdividing its range into a set of equally-sized bins. This representation of the Bayesian model allows the conditional probability of each discrete value (or range of values) of a feature to be stored in the Cytoscape network as an edge table attribute. However, as Figure 3 demonstrates, this structure is incredibly verbose and requires multiple nodes to separate the bins for a single feature. We instead provide a simpler way for users to design a Bayesian template (an example is shown in Figure 2).\n\nUsers may define custom Bayesian templates using Cytoscape’s native graph-building or importing tools. These templates must use a specialized syntax for naming nodes, wherein each node other than the root (labeled ‘Root’) specifies a feature. The syntax for a feature is as follows:\n\n\"Statistic\" is an optional prefix that indicates an operation to perform on the values of a feature (such as taking the max, mean, or median). A statistic must be applied to a feature that produces a list of values rather than a singular value. Applying the statistic results in a single floating-point value, which may then be assigned to the appropriate bin. The available statistics include: Mean, Median, Count, and ordinals such as 1st, 2nd, 3rd, etc.\n\nFor example, Figure 2 illustrates a simple Bayesian network with two features that are conditionally independent given the root node. The feature labeled \"density\" may have continuous values in the range (0, 1], divided equally among 4 bins. In the second feature, \"degree\" represents the number of neighbors of each node in a candidate cluster. Applying the \"max\" statistic produces the highest degree among all the nodes in a cluster. The range of this feature is (0, n], where n is the highest degree of any node in the positive training data.\n\nOnce its features are defined, the template is trained using a set of user-provided positive training complexes and a set of randomly-generated negative training clusters. First, two nearly-identical representations of the Bayesian template are created, with the root node of the first labeled ‘Root Cluster’ and the root node of the second labeled ‘Root Non-Cluster’. A separate node is then created for each bin of each feature, resulting two networks whose respective sizes are, identically, the sum of the number of bins for each feature plus the root node. Figure 3 illustrates the two networks produced after training the template in Figure 2.\n\nThe positive Bayesian network (‘Root Cluster’) is trained using the positive complexes provided by the user, and the negative Bayesian Network (‘Root Non-Cluster’) is trained using the randomly-generated negative protein clusters. In both training procedures, the program records the frequency at which the values representing each node in the Bayesian network are encountered among the training clusters; this translates to the frequency for each feature bin.\n\nBy the end of training, the node frequencies are used to encode the conditional probabilities, P(fb|RootCluster) and P(fb|RootNon − Cluster) for bin b of feature f, as follows:\n\nWe may then use Equation 1 to produce a log-likelihood score for a candidate cluster based on the conditional probability of each feature bin in both the positive and negative Bayesian networks. Conditional probabilities on the root node are maintained as properties of the edges in the returned Cytoscape network, which represents both halves of the trained Bayesian model; in doing so, trained models may be saved and recycled for later use in Cytoscape session files.\n\nThe search phase returns a set of predicted complexes and their associated likelihood scores, which may be visualized and analyzed independently using Cytoscape tools. SCODE also provides a tool for evaluating discovered complexes against a user-provided set of known complexes in the network, allowing the program to quantify its overall performance.\n\nTo perform the evaluation, the user must supply a file containing a list of complexes known to exist in the PPI graph. These complexes may include the positive examples used to train the Bayesian network, but they will be filtered out when calculating the evaluation score.\n\nThe evaluation provides two metrics: recall and precision. Recall measures the ability of the program to discover complexes from the known set, while precision measures the accuracy of the discovered complexes.\n\nWhen comparing a predicted complex against a known complex,\n\n• Let A be the number of proteins only in the predicted complex\n\n• Let B be the number of proteins only in the known complex\n\n• Let C be the number of proteins in both the predicted and the known complex\n\nA predicted complex is said to have identified a known complex3 if\n\nHere p is a hyperparameter specified by the user. Recall and precision are calculated as follows:\n\n\n\nSCODE is organized among three packages. The statistic package contains implementations of the abstract class ’Statistic’, which specifies methods for operating on a list of feature values and calculating the overall range of those values for binning. A second package, feature, contains implementations of the abstract FeatureSet class. FeatureSet specifies methods for getting feature values, applying statistics to them, and returning the bins for a complex. The list of statistics that is applied to a feature is stored as a protected member of the FeatureSet class.\n\nThe third package contains the remaining code for gathering input and executing the search, scoring, and training tasks. The entry point to SearchTask and TrainingTask creation is in SupervisedComplexTaskFactory. TrainingTask initiates the process of loading the appropriate template network or trained model network, which is then used to create the internal representation of the Bayesian network in the SupervisedModel class. SupervisedModel represents both the positive and negative Bayesian graphs via the Graph class, which also specifies methods for training the graph and scoring candidate complexes during search.\n\nFrom the initial searchTask, an IsaSearch object is created to divide the starting nodes among separate threads and to begin searching on each one. The core search algorithm is located in the SeedSearch class, where candidate complexes are expanded and scored. Complexes are represented internally as Cluster objects and returned as CySubNetworks once the search has been completed.\n\nTaking advantage of its new modular architecture, SCODE requires Cytoscape version 3.2 or above to operate. Upon launch, SCODE accepts user input and displays a summary of the results from the SCODE tab in the Control Panel. Mirroring its pipeline, SCODE’s interface is divided into subsections for each of the search, scoring, training, and evaluation tasks.\n\nThe user may adjust the following input parameters for the search stage:\n\n• Variation of Simulated Annealing : The program offers three variants of the Iterative Simulated Annealing algorithm: ISA, Sorted-Neighbor ISA, and Greedy ISA.\n\nISA : Fastest performance but tends to produce low-scoring complexes.\n\nGreedy ISA : Slower than ISA, but tends to produce more, larger, higher-scoring candidate complexes.\n\nSorted-Neighbor ISA : Slower than ISA and sometimes slower than Greedy ISA, depending on the density of the PPI graph. Tends to produce higher scoring complexes.\n\n• Search Limit : Specifies the maximum number of iterations of the search.\n\n• Initial Temperature : Sets the starting temperature of the search. Higher temperature increases the likelihood that a protein will be added to the complex.\n\n• Temperature Scaling Factor : The rate at which the temperature decreases over time.\n\n• Overlap Limit : The maximum proportion of nodes that two distinct complexes may share.\n\n• Use Seeds From File : Optional; the user may select an external file containing a list of seed nodes from which to begin the search.\n\n• Number of Random Seeds : Allows the program to randomly select the specified number of seed nodes.\n\n• Number of Results to Display : The number specified here will be the number of results visible at the end of the search.\n\nThe scoring section accepts the following parameters:\n\n• Minimum Complex Score : When the results are shown to the user, only candidate complexes with scores equal to or greater than this input will be displayed.\n\nThe training section accepts the following parameters:\n\n• Cluster Probability Prior : The prior probability, P(x1), that a group of proteins forms a complex.\n\n• Generate Negative Examples : Specifies the number of negative training examples that the program will randomly generate.\n\n• Ignore Missing Nodes : During training, if one of the proteins in a positive training example cannot be found in the PPI network, selecting this option will allow the program to disregard the training example.\n\nThe training section takes additional input parameters relating to the Bayesian network. The first set of parameters specifies the Bayesian network itself. The user may either use a default model provided by the application, or a custom Bayesian network that has been constructed using Cytoscape. The custom network may be either trained or untrained; in the latter case, the user must provide an input file containing a set of known complexes that will be used to train the model. Trained models can be saved and loaded from Cytoscape session files, which store networks for later use.\n\nSCODE includes a number of predefined features for building Bayesian templates, enumerated below. The features of the built-in Bayesian template are provided in Table 2, but are not exhaustive. In the following descriptions, G indicates a PPI graph with edges, E, and nodes, N:\n\n• Complex Size : Takes the length of a list of complexes of size |N|.\n\n• Clustering Coefficient : How many triangles contain the node n as a vertex. Calculated using the equation CCn = 2ek/(kn * (kn − 1)), where kn is the number of neighbors of n and ek is the number of edges between those neighbors.\n\n• Degree : For a node in N, the number of edges connected to that node.\n\n• Degree Correlation : For a node in N, the average number of neighbors among its adjacent nodes.\n\n• Density : The ratio of the number of observed edges to the number of possible edges in G: D = |E|/|Ep|.\n\n• Density at Cutoff : The same as above, once edges with a weight below the cutoff are removed from the set of observed edges.\n\n• Edge Weight : A measure of the strength of the interaction between a pair of nodes. Must be provided as an edge table column in the PPI network.\n\n• Edge Table Column : Produces a vector of values from a Cytoscape graph table’s column for each edge in the graph.\n\n• Edge Table Correlation : Produces a set of vectors from a set of columns in a Cytoscape table and returns a list containing the correlation coefficients for each pair of vectors.\n\n• Node Table Column : The same as the edge table feature, but applied to each node.\n\n• Node Table Correlation : The same as the edge table correlation feature, but applied to nodes.\n\n• Singular Values : The singular values for each graph correspond to the graph’s shape. For instance, A linear graph’s values will consistently differ from those of a clique.\n\n• Topological Coefficient : For a node n, the topological coefficient measures the connectivity of n’s neighbors, where f (a, b) is the number of neighbors in common between neighbors a and b, while kn is the number of neighbors: TCn = avg(f (a, b))/kn\n\n\nUse cases\n\nWe demonstrate the app using training/evaluation datasets constructed from two sources: the CYC2008 catalogue of manually curated protein complexes4, and the TAP06 dataset of complexes screened using affinity purification and mass spectrometry5, which were each filtered to a random sampling of 50 complexes with 4–6 nodes. Both files are supplied in a tab-separated format.\n\nThe graph on which search is performed features 396 proteins (those featured among the CYC2008 and TAP06 complexes) and 5141 edges. The set of interactions in this graph is generated from the STRING database of known and predicted protein-protein interactions6. The graph is also in tab-separated format and must be loaded into Cytoscape as a network; the network may then be saved in a session file (.cys) for later use. All of the data used in this demonstration can be found online and is provided below under ‘Data Availability’.\n\nAt the end of the search, discovered complexes will be returned as Cytoscape networks under the ‘Network’ tab in the Control Panel.\n\nFor this demonstration, we employ the parameters shown in Figure 4. The number of random seeds matches the number of nodes in the graph. This number, to some degree, dictates the likelihood of identifying complexes of good quality. With a low starting seed count in a large PPI graph, the probability of beginning the search from a protein that is a member of a \"true\" complex is low. For the purposes of our evaluation, we provide the ideal conditions for the search to return true complexes. This includes using Greedy ISA so that all nodes neighboring a complex are scored before selecting one for expansion.\n\nWe choose to employ the built-in Bayesian template for training, with features shown in Table 1. Since the PPI graph is under 500 nodes, we use a prior probability of 0.5 for clusters, and generate 50 negative examples. Two rounds of training and search are performed, the first using the abridged set of CYC2008 complexes for training and the TAP06 complexes for evaluation, and the second using the reverse (with all other parameters held constant). The minimum complex score is set to 0.\n\nOnce the search is complete, the top M results (specified under the search parameters) are displayed under the ‘Network’ tab in the Cytoscape Control Panel. All protein clusters produced by the search, including but not limited to those displayed, are considered during evaluation. The Evaluation section appears directly below the Scoring section after ‘Analyze Network’ is clicked (Figure 6).\n\nWe supply p=0.5 for Equation 3, such that all predicted complexes with a majority of proteins from a known testing complex are said to have \"recovered\" that complex. After clicking \"Evaluate Results\", the evaluation scores for recall and precision appear in a dialog (Figure 7).\n\nEvaluation scores using CYC2008 to train and TAP06 to test (a) or TAP06 to train and CYC2008 to test (b).\n\n\nSummary\n\nSCODE expands the detection of protein complexes in weighted PPI networks by applying a supervised learning algorithm with a set of known training complexes. Users may discover topologically non-traditional protein complexes by leveraging more information about the features of its PPI graph. The Bayesian network encodes the desired characteristics of complexes beyond density and is used to score the likelihood that a candidate discovered during a search of the PPI network represents a complex. Each version of the ISA search heuristic discovers complexes of varying quality and size, in accordance with the degree to which the algorithm exhausts the search space.\n\n\nData availability\n\nF1000Research: Dataset 1. Demo Using CYC2008 and TAP06 Complexes, 10.5256/f1000research.9184.d1288788\n\n\nSoftware availability\n\n1. Software available from:\n\nhttp://apps.cytoscape.org/apps/scode\n\n2. Latest source code: https://github.com/DataFusion4NetBio/Paper16-SCODE\n\n3. Archived source code as at time of publication: http://dx.doi.org/10.5281/zenodo.571639\n\n4. Software license: MIT License (https://opensource.org/licenses/MIT)",
"appendix": "Author contributions\n\n\n\nYQ conceived of, funded, and supervised the project. NJ and SM implemented the app. SM, NJ, and YQ authored this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the National Science Foundation under Grant No. 1453580. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Leonard Ramsey of the University of Virginia for his feedback and assistance in improving the functionality and supporting documentation of the app.\n\n\nReferences\n\nIdeker T, Sharan R: Protein networks in disease. Genome Res. 2008; 18(4): 644–652. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBader GD, Hogue CW: An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003; 4: 2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQi Y, Balem F, Faloutsos C, et al.: Protein complex identification by supervised graph local clustering. Bioinformatics. 2008; 24(13): i250–i268. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPu S, Wong J, Turner B, et al.: Up-to-date catalogues of yeast protein complexes. Nucleic Acids Res. 2009; 37(3): 825–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGavin AC, Aloy P, Grandi P, et al.: Proteome survey reveals modularity of the yeast cell machinery. Nature. 2006; 440(7084): 631–6. PubMed Abstract | Publisher Full Text\n\nSzklarczyk D, Franceschini A, Wyder S, et al.: STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015; 43(Database issue): D447–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMewes HW, Frishman D, Gruber C, et al.: MIPS: a database for genomes and protein sequences. Nucleic Acids Res. 2000; 28(1): 37–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMohamed S, Janus N, Qi Y: Dataset 1 in: SCODE: A Cytoscape app for supervised complex detection in protein-protein interaction graphs. F1000Research. 2016. Data Source\n\nMohamed S, Janus N, Qi Y, et al.: SCODE: Supervised Complex Detection. Zenodo. 2016. Data Source"
}
|
[
{
"id": "16428",
"date": "28 Nov 2016",
"name": "Narayanaswamy Srinivasan",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors employ a supervised learning algorithm for prediction of protein complexes from protein-protein interaction networks. This manuscript requires considerable work. I am listing what I think as most crucial points for authors to work on:\nThe input to the proposed development is protein-protein interaction network. This information usually comes from laboratory studies. While there is always an element of accuracy and completeness associated with the data, once a network is constructed what are the protein assemblies within the network is intuitively obvious. Why do we need a development such as the one proposed in this manuscript? Where is the question of “prediction” of assemblies? – the interaction information is already firmly coded in the network and information on assemblies is part of construction of the network. This is a fundamental and serious problem authors must address.\n\nAuthors have come up with a computational development. But there is no assessment on how well the method works.\n\nThere is no application shown using a protein-protein interaction network.\n\nManuscript reads too technical with too much jargon.\n\nFigures could be lot more informative. For example the two parts of Figure 3 are extremely similar and the message from the figure is not clear.",
"responses": []
},
{
"id": "20480",
"date": "04 Apr 2017",
"name": "T.M. Murali",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n4th April 2017: This referee report was mistakenly published with a Not Approved status - we have now corrected this to the intended Approved with Reservations.\nThe manuscript present a Cytoscapp app called SCODE for supervised detection of protein complexes. In principle, the motivation of the work is strong: learning the features of known complexes in a protein interaction network is a viable strategy to compute new complexes. However, the manuscript requires a significant amount of revision.\nThe authors motivate the usefulness of supervised learning of complexes by saying that the problem is often reduced to clique detection. They are considerably simplifying the literature in making this claim, since there are literally dozens of algorithms for computing protein complexes in protein networks and they cite only one publication (the MCODE paper) to support the claim.\nThe authors claim that \"Real-valued data indicate that a variety of other features, including but not limited to edge density, are typical of complexes among PPI networks.\" In support, they cite an earlier paper from 2008 (citation 5). In citation 5, the authors of that paper show a few example of complexes that are not-cliques. It is possible that algorithms developed since 2008 may compute such subgraphs. This manuscript will be improved if the authors provide some more recent and updated data to support the claim.\nIt is unclear how the SCODE algorithm differs from the method published in citation 5. The manuscript should clarify if the two methods differ in any essential features or if this manuscript focuses on presenting a Cytoscape app that implements the method from citation 5.\nIn either case, the precision of SCODE is very low (around 0.1) when the recall is around 0.5. The manuscript will be much improved by a quantitative comparison to other algorithms (supervised or not) in terms of its precision and recall.\nA difficulty that a user of the SCODE app will find is that it has a surfeit of parameters. When faced with so many choices, a user is likely to settle for the defaults without knowledge of whether these values will lead to good results or not. The usability of the app can be dramatically improved by decreasing the number of parameters and providing some guidelines to the user on how to tune them for a particular dataset. In the section on \"Search parameters\", the authors do state \"For the purposes of our evaluation, we provide the ideal conditions for the search to return true complexes.\" They should explain how they determined that these parameters provide ideal conditions to find true complexes.\n\nThe descriptions of some features are confusing and unclear. For example, \"Complex size\" is \"Takes the length of a list of complexes of size |N|.\" Do the authors mean that the feature is for a complex c, the number of proteins in c? They use this type of language for features such as degree and clustering coefficient. In the definition of density, what is Ep? What are the values in \"Edge Table Column\"? What are the singular values of a graph? How do the authors compute them? What is their biological meaning?\nMore generally, some features are defined for nodes, some for edges, some for pairs of nodes, some for the entire graph. It will be helpful to group them by this categorization and explain how the different types of features fit into the training framework.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1699
|
https://f1000research.com/articles/5-1224/v1
|
06 Jun 16
|
{
"type": "Opinion Article",
"title": "Where are the female science professors? A personal perspective",
"authors": [
"Shina Caroline Lynn Kamerlin"
],
"abstract": "The first woman to earn a Professorship at a University in Europe was Laura Maria Caterina Bassi, who earned a professorship in physics at the University of Bologna in 1732. Almost 300 years and three waves of feminism later, in 2016, women typically still only comprise 20% (or less) of the number of full professors in Europe. This opinion article will discuss the experiences of being a female academic today and the factors contributing to the academic gender gap from the perspective of a “young” natural scientist, as well as providing constructive suggestions for strategies to empower women in the academic world.",
"keywords": [
"Women in science",
"Implicit bias",
"Academic gender inequality",
"Matilda effect",
"Empowering female academics"
],
"content": "Introduction\n\nAs women occupy an increasing number of prominent roles in society, it is easy for us to forget just how recent many advances in women’s rights that we currently take for granted actually are. For example, women have only held the right to vote for about 100 years or less in most European countries1. Women have only been allowed to attend European Universities (and initially often only as auditors) since the late 1800s2. In Sweden, before 1859, women did not even have the right to be college teachers3. In such an environment, it is perhaps unsurprising that, when Lise Meitner headed to Germany with the hope of working at the University of Berlin in the early 1900s, women were not even allowed on the premises4. Almost half a century later, when Rosalind Franklin was a research fellow at King’s College London, women were still barred from even entering the senior common room5, effectively cutting them off from college life.\n\nDespite such hostile-to-women environments, female pioneers in science and technology have made countless contributions to science, including developing the first published computer algorithm (Ada Lovelace Byron)6, developing the technique of X-ray crystallography (Dorothy Hodgkins)7, taking the first X-ray diffraction image of DNA (Rosalind Franklin)5, giving us unprecedented insight into the world of chimpanzees, which also helped us better understand ourselves (Jane Goodall)8, co-discovering nuclear fission (Lise Meitner)4 and co-inventing a frequency hopping communication system that is the basis for modern day WiFi technology (Hedy Lamarr)9, to name just a few examples. However, even with the seminal contributions of such women to natural sciences, technology, social sciences and the humanities, we are very far from achieving anything near gender equity at the senior levels of academia. I will explore herein the causes of this inequality and some of the barriers facing women’s progression in the academic world. I will also discuss briefly some of the work I have personally engaged in to fight this inequality, as well as providing suggestions for how we can all contribute to create a more equal working environment.\n\nThe focus of this opinion article will primarily be on natural sciences and the academic ladder in Sweden, simply because these are the worlds I know best. I strongly emphasize that this is in no way meant to be a value judgment on the Swedish system with respect to other European academic systems, but rather I use Sweden as my example only based on familiarity. Additionally, clearly gender is not the only form of inequality in the academic world, and arguably many other forms of inequality are even more aggravated. The fact that I do not specifically discuss these here is not due to lack of interest, but rather due to the finite scope of this article. Despite these limitations, I believe that many of the discipline- and country-specific challenges we face have universal underlying roots, and therefore that my overall observations are independent of discipline and country.\n\n\nExamining the current state of gender (in)equality in academia\n\nIn order to set the scene, I would like to start by discussing some statistics of the participation of women in Swedish academia. By all measures, Sweden is a pioneering country in terms of gender equality parameters. For example, Sweden consistently ranks very highly on the World Economic Forum’s Global Gender Gap report, coming in at 4th place in 2015, behind only three other Nordic countries (Norway, Finland and Iceland)10. Having achieved such a standing in the Global Gender Gap rankings is something I believe the Nordic countries should justifiably be very proud of. However, despite overall equality in these societies, once one moves to the Academic world, the situation changes rapidly. That is, as can be seen from the European Commission’s 2012 She figures for the % of women in Grade A and B positions in Europe in 2012 (Figure 111), not only does Sweden no longer occupy the top positions for female participation in Grade A positions, it barely makes the European Union average, coming in at 13th place. I will return to discuss these statistics later in this opinion article; however, it is worth noting that this comparison also throws up an unexpected surprise: arguably among the most socially conservative countries in Europe, Turkey also has the among the largest number of female professors among European countries (ranking first in 2005 in data by ref. 12), whereas countries such as Germany, Austria, the Netherlands and Denmark, in fact all come in at the bottom half of the list. Note here that a Grade A position is defined by the Commission as the single highest post at which research is normally conducted (in this case a Professorial position), and a Grade B position comprises researchers that are more qualified than newly minted PhD holders, but not as senior as those in Grade A.\n\nData obtained from the 2012 European Commission She figures for gender in research and innovation11. For description of the source data see ref. 11.\n\nThere are also a number of stereotypes about women’s involvement in academia that I believe need to be readdressed. For example, when discussions are held about how gender parity could be achieved in academia, it is often asked “why are women leaving?”13. Therefore, I would like to present two more sets of statistics, shown in Figure 2 and Figure 3. Figure 2 shows the percentage of women in different academic career stages after having started a PhD, data from 2013 and averaged over all disciplines14. From this figure, it can be seen that like many other countries, Sweden has a vertical gender balance in academia, such that women comprise the majority (55%) of university entrants, just slightly under the majority of doctoral and postdoctoral candidates, and even senior lecturers/associate professors (46, 42 and 42% respectively), and yet there is a very sharp decline such that this number drops to only 20% of full Professors. This occurs across all disciplines (Table 1)14, and worryingly even in otherwise heavily female dominated fields such as pharmacology and veterinary medicine. However, importantly, the statistics show that at both undergraduate and doctoral levels, irrespective of the percentages of entrants to these degree programs, women are slightly more likely to stick it out and actually be awarded a degree.\n\nData obtained from ref. 14.\n\nFinally, as shown in Figure 3, if one compares the percentage of women and the number of men who have left academia five years after the award of a doctoral degree14, one sees that in all disciplines except social sciences, contrary to stereotypes, more men are likely to leave academia than women, although this is then not reflected at the highest levels, for example in the number of full professors. This then raises some really crucial questions: where did all the women go? What barriers are facing women in academia today? And how can we empower more women to lead and excel in the academic world?\n\nData obtained from ref. 14, based on the 1993 cohort of doctoral students.\n\nGender studies is a broad field, and many hypotheses have been put forward to rationalize the lack of women in the academic world. Ceci and Williams have summarized these15, providing three general broad arguments that are put forward to explain the dearth of women in academia:\n\n1. The fraction of women who have the native intellectual capacity to do science, particularly at the highest levels, is much smaller than the fraction of men. I should note that I personally find this argument deeply offensive, but it is lamentably a not uncommonly held belief, as was demonstrated in its most high profile example in 2005, when then Harvard President Larry Summers claimed at a conference that the barriers to women’s advancement in academia have been removed, and that the underrepresentation of women at elite universities may stem from “innate” biological differences in ability between the genders16.\n\n2. An inherent lack of interest among women in the hard sciences and engineering.\n\n3. Societal and cultural biases that push women out of the pipeline and lead to the devaluation of those that remain.\n\nI will proceed to systematically discuss the main barriers I observe through both my own experiences as a female academic and from discussion with my colleagues below.\n\n\n\nIn most European countries, there have been major advances and improvements in mechanisms to allow women to balance career and family obligations. Paid maternity leave, extensions on grants that take into account childcare responsibilities, availability of time off to care for children when they are sick are among only a few of these advances. However, the problem still remains that the crucial formative early years that determine a young scientist’s future career trajectory also coincide in age with the years in which many young scientists need to start seriously considering their family plans. Clearly, balancing the two is not easy, particularly in an environment where hyper-competition is now the norm to attain coveted grants and permanent faculty positions17. Gender studies of course take this challenge into account, and it is a large research area, where the literature can take on scathing titles such as “Career progress relative to opportunity: How many papers is a baby ‘worth’?”18, “How much do children really cost?”19; “Balancing work-family life in academia: The power of time”20, and “Pinstripes and breast-pumps: Navigating the tenure-motherhood track”21. This issue has also been taken up at great length in recent years in editorials and opinion pieces in leading newspapers and magazines, such as Slate Magazine22, the New York Times23, the Chronicle of Higher Education24, and the Atlantic25.\n\nIf one were to summarize the viewpoint of these latter publications on balancing children and an academic career track for female academics, they could be distilled into one simple word: “Don’t”. Additional arguments put forward include that while becoming a parent is not necessarily as bad for a man, having children is a career killer for a woman. This is in practice not always the case, of course, and while children clearly pose a particular challenge, many women have gone on to have successful academic careers while being mothers. Such women, in turn, can make a major contribution as role models for younger colleagues. Additionally, as I will discuss further in this section, ongoing biases against women in academia in my opinion pose a much larger problem than balancing family issues, at least in Sweden. However, it does remain a challenge, as childbirth is often a point at which many women either decide or are forced to leave an academic career trajectory, due to competing personal and professional obligations. For example, a 2009 survey of University of California postdoctoral fellows (Figure 4) showed that those who already had or were considering having children were more likely to also consider leaving research26. Additionally, the penalties on women who decide to try to have children and not leave research are quite severe. Well-document coping strategies include: waiting until tenure to have children, not having children at all, timing children around the academic calendar, moving to part-time work, increasing research collaborations (presumably to hide “lost time” due to childcare responsibilities), sleeping less, sacrificing personal lives, and moving to “second-tier” institutions18. Clearly, although these strategies are employed as a means to cope, they will have a highly detrimental effect on a female academic’s scientific productivity and career progress.\n\nResults from a 2009 survey of postdoctoral fellows at the University of California, based on data presented in ref. 26.\n\nAdditionally, while there has been massive progress in provisions for helping women balance careers and family life, in particular in terms of maternity leave, these are a mixed blessing. Central European countries such as Sweden, Denmark, the Netherlands and Germany have amongst the longest paid parental leave provisions in the world, at 48027, 36428, 11229 and 9830 days parental leave respectively. One would assume from this that these countries also therefore provide excellent opportunities for women to integrate into the workplace, yet as shown in Figure 1, in particular Denmark, the Netherlands and Germany are among Europe’s poorest performing countries in terms of female integration into the academic world at senior (Grade A) positions. Additionally, Turkey, with the same length maternity leave as the Netherlands (112 days)31, also has among the highest percentage of female professors among European countries.\n\nClearly, therefore, the link between length of maternity leave and professional success in academia is not as straightforward, and other factors including childcare provisions and societal attitudes play a major role. For example, of the 480 days parental leave in Sweden, 60 days are reserved explicitly for the father, and parents are strongly encouraged to split time equally between them. This would of course have a benefit of distributing the burden of childcare, although there have also been criticisms of the fact that “time” is not divided equally among the genders and the more stressful of childcare duties such as putting uncooperative children to sleep, getting up in the middle of the night to tend to their needs, and the frantic rush to get them out the door in the morning, is more often than not taken on by the female member of the household32. Additionally, even with encouragement to try to split time in Sweden, according to 2012 statistics, women still took 76% of parental leave days with men only taking 24%, and only 13% of parents share parental leave days equally33. Therefore, Sweden is long from splitting parental responsibilities equally between both parents34, and this inequality has been directly linked to both an increased gender-based wage gap and also to hardening the glass ceiling for women in Sweden35. That is, ironically, despite the numerous measures to promote gender equality in the Nordic countries, women in the Nordic countries are actually less likely to reach top leadership positions, compared to, for example, the United States, which has fairly minimal regulations with respect to childcare and maternity leave36. This is due to a combination of many factors: actually taking the extended maternity leave options offered can lead to women becoming “mommy tracked”, while simultaneously becoming rusty on important career skills and social contacts36, which impairs opportunities for further career development. Therefore, while work-family commitments are not the only barrier to the empowerment of women in the academic world, even in 2016, they clearly form a major part of the problem.\n\nIn 1968, Merton coined the term the “Matthew” effect, to describe over-recognition of those at the top of the scientific elite, which can extend to even credit misallocation to already well-known scientists37. Following from this, in 1993, Rossiter borrowed this concept to coin the term “Matilda” effect38, which refers to the systematic under-recognition of the contributions of female scientists. The question is, therefore, whether such a “Matilda effect” actually exists in science. While I would really like to be able to say no, unfortunately, there is a large amount of qualitative and quantitative evidence pointing to the contrary.\n\nThe biggest challenge with the Matilda effect, i.e. systematic bias and discrimination against the contributions of women, is that its roots start at a very early age. For example, in 2007, Steinke and coworkers performed the “Draw-a-Scientist” test39. This was essentially a sociological experiment, to get 304 seventh-grade students, to draw what they think a scientist should look like. A summary of the characteristics attributed to male vs. female scientists are summarized in Table 2. From the statistics it can be seen that already in the seventh grade, children are heavily influenced by media stereotypes, with the vast majority of children believing that a scientist is a man, in a lab coat and glasses, and 42.4% also assumed that scientists are stern and do not smile. While this may seem whimsical in itself, the implications are severe, because it suggests that already at a young age, children have a distorted image of what it means to be a “scientist” – and therefore a distorted image of their own ability to be an excellent scientist.\n\nBased on data presented in 39.\n\nUnfortunately, the bias that was already being observed in these young middle-schoolers does not go away, but rather is consolidated as the children grow up progress through the academic ranks. For example, in 2010, Amy Bug from Swarthmore College performed another sociological experiment, in which 126 students had to watch 4 ten-minute lectures given by two male and two female physics professors40. The students then had to evaluate both the lecture, and the professor’s knowledge ability. What Bug observed was that, on average, female students gave slightly higher marks to the women than to the men, but that this was more than compensated for in the fact that male students on average gave massively higher marks to men than to women. In addition, neither group was aware of the fact that their professors were paid actors, reading from exactly the same script, with no prior background in physics!\n\nTaking this one step further, when it comes to recruitment of undergraduate lab assistants, Moss-Racusin and coworkers41 performed a randomized double blind study in which a broad, US-wide sample of science faculties (n=127) received a hypothetical application pack for recruitment to a position as an undergraduate laboratory assistant. The materials were randomly assigned either a male (n=63) or female (n=64) name. All other parameters were identical. Faculty members were then asked to rate students’ competence, hirability, and the salary and mentoring they would offer the student. The results of this are shown in Figure 5. Critically, all faculties believed that students would see the feedback. From this figure, it can be seen that not only were “male” lab assistants routinely deemed to be more hirable, competent and worthy of mentoring, the salary gap for applicants with exactly the same CV was in excess of $3000/year. Additionally, both male and female faculty members judged the female student as less competent, and less worthy of being hired than an identical male applicant, and also offered her less salary and mentorship. This faculty member bias was observed to be independent of gender, scientific discipline, age and tenure status. Female and male faculty members are equally biased. While this may in itself again seem to be just a trivialized local study, clearly such subconscious bias can in turn translate into large real world disadvantages in the judgment and treatment of female students. This in turn raises concern about the extent to which negative pre-doctoral experiences may shape women’s subsequent career decisions41.\n\nCollapsed results, independent of gender of evaluator, showing (A) competence, hirability and mentoring scores (assessed on a scale from 1 to 7), and (B) offered salaries, for male and female students. Based on data presented in ref. 41. Error bars represent standard deviation over the two genders. For methodological details and raw data, see ref. 41.\n\nFinally, in a by now quite famous study on sexism in Swedish peer review42, the authors explored the discrepancy between the fact that in 1997, women were awarded 44% of Swedish biomedical PhDs, but held only 25% of postdoctoral and 7% of professorial positions (a statistic that Figure 1 shows has fortunately almost tripled in under 20 years). Additionally, the success rates of women applying to prestigious medical research council (MRC) postdoctoral fellowships was only half that of male applicants. The reasons put forward to justify this were variants on the theme presented at the start of this section, that women are “less productive”, “less motivated”, “less career oriented”, “less …”. Unconvinced, the authors decided to explore whether reviewers can truly do a “gender-free” evaluation. At the time, the postdoctoral fellowship applications comprised of a Curriculum Vitae, publications list and proposal. These were then reviewed by five people from one of 11 topical evaluation committees. Each reviewer awarded the applicant a score of between 0 and 4 each for scientific competence, relevance of the research proposal, and the quality of the methodology. The scores were then multiplied to give a product score between 0 and 64, and averaged over all five reviewers to give a final score to the applicants, with candidates being ranked according to the final score received. Under Sweden’s Freedom of the Press Act, the authors were allowed to see the MRC evaluations by court order. From this, they observed that female applicants had lower scores on all three evaluation points, and that these were, on average, 0.25 lower for competence, 0.17 lower for methodology and 0.13 lower for the quality of the research proposal. The multiplicative nature of the different criteria then led to substantially lower overall final scores.\n\nWhat stood out from this assessment was the fact that the female candidates appeared to be deemed particularly deficient in scientific competence compared to their male counterparts. Since assessment of scientific competence is normally related to the number and quality of the applicant’s scientific publications, the authors wondered if the female applicants are really less productive than the male ones. To assess this, the authors constructed a model, in which each applicant was given a “mean competence score”, as a function of scientific productivity, measured as total impact. On top of this, the authors used multiple regression analysis to correct for external factors such as nepotism, university affiliation, and connections to members of the evaluation committee. Once completed, the authors observed that, even after correction, the female applicants needed to, on average, be 2.5 times as productive as man for the same competence score, with a worrying trend that the higher impact the applicant, the higher a male applicant’s contributions were scored compared to their female counterpart (an extreme incarnation of the Matilda effect). In concrete terms, this translates to three extra papers in Nature or Science, or 20 extra paper in a journal with an impact factor of 3, and with such expectations it would hardly be surprising that fewer women manage to achieve academic career success than men.\n\nClearly, the Swedish Research Council responded strongly to these observations, and put systems in place to improve the peer review process and promote applicants receiving equal treatment irrespectively of gender. While things have improved dramatically since then, in particular with women receiving 35% of grants awarded by the Swedish Research Council for 201543, there are still clear areas that need addressing before a truly “gender free” peer review process can be achieved44. Additionally, clearly gender bias in peer review is far from a uniquely Swedish problem, and a 2007 meta-analysis of 21 such studies demonstrated that, on average, male applicants have a 7% greater change of obtaining research funding than female applicants, which can be quite a dramatic difference at a time when grant success rates are going into the single digits45. This is problematic, in particular in light of the fact that in such a low-success environment, even small biases can have major negative impact46 (creating a “mountain of feathers”). Many other examples of the Matilda effect have also been observed, in selecting women for conference presentations47,48, assessing publication quality and citations49,50, recruitment and tenure processes51,52, and even in recent arguments that elite male faculty members in the life sciences employ fewer women than men to their labs, thus creating an unbalanced career start for young female scientists53, or that papers in which the lead author is a man are more likely to get cited than corresponding work led by a woman49. Therefore, unfortunately, the Matilda effect is alive and well in science, and one of the biggest current barriers to the true empowerment of women in the academic world.\n\n\nFocus 2016: how can we empower women in the academic world?\n\nHaving explored some of the major barriers to women in the academic ladder, I would like to focus on constructive examples of how these barriers can be removed, in order to achieve greater gender equity in academia.\n\n\n\nTo open this section, I would like to briefly come back to the example of Turkey, which is leading in Europe in percentage of women in academic positions12. In some cases where women are highly represented in senior academic positions (Figure 1), it has been argued that this is in part because of the willingness of women to take more insecure and poorly paid career trajectories, and that academia in general is considered a less prestigious career path, thus increasing female representation in this sector (see for example ref. 54, and references cited therein). However, in this respect, Turkey’s example is therefore worth highlighting, because the high female representation in Turkey is not by accident, but rather a result of concrete policies over a longer period of time. In particular, Healey and coworkers argue that the higher representation of Turkish women in academia can be brought down to five key features of the Turkish system55:\n\n1. The existence of historical long-term state-driven ideology, promoting the participation of women in the Turkish academic labor force.\n\n2. The fact that, in general, academia has been considered a “female appropriate” career choice, resulting in little gender disparity among university graduates (even in the sciences).\n\n3. The existence of significant university expansion in the 1990s, which created demand for both male and female professors.\n\n4. The existence of a comparatively transparent employment and promotion system.\n\n5. The reliance of female faculty on domestic help, making it easier to balance family and professional commitments.\n\nClearly, when these factors are brought together, they lead to a holistic picture that is productive for the empowerment of women in the academic world, and show that even though there are many structural and practical problems that still need addressing, nevertheless, the barriers facing the empowerment of women in the academic world are surmountable ones, if we are only willing to take them on.\n\nA few years ago, in discussion with a postdoctoral scholar in a colleague’s research group about her career prospects, she asked me why she should even remain in academia, when there are no women. This remained with me and is in part the reason for why I have taken a lot of the mentorship work I discuss in the concluding section. Young women need strong role models: it is important for successful women in academia to be one. This can be achieved in many ways. Mentoring of junior colleagues is particularly important, as is encouraging them to actually apply for grants, fellowships, faculty positions and promotions. In my work mentoring junior colleagues, I often hear my mentees insist that they are not yet ready to do so, what if they only had just a few more papers, the call for appointment is not really in their field, and maybe they can apply the next year. This creates a problem, because it means that many women don’t think they are good enough, and don’t get on the academic ladder in the first place. This can be addressed through greater mentorship opportunities, as well as raising the visibility of women who do exist in academia.\n\nTo partly address this issue, I have together with the Young Academy of Europe and Uppsala SciLifeLab organized a one-day symposium at Uppsala University, with 12 outstanding speakers from disciplines across natural sciences and technology, and four similarly prestigious session chairs56. This was tremendously successful, with the participation of 166 delegates from 12 different countries and four continents. Additionally, the University of Southern California Women in Science and Engineering (WiSE) program have compiled a database of women in theoretical/computational chemistry, material science and biochemistry57. In these fields, women provide only a smaller percentage of total faculty in any given department, and can therefore easily be lost in the crowd at individual institutions. However, this database highlights the fact that globally, there are several hundred examples of women working in these research areas, and provides for example a quick reference list of outstanding women one can refer to when putting together seminar series, conference speaker lists, and similar activities. Such lists can provide a quick reference point for conference organizers who want to ensure a more equal gender distribution when planning meetings and symposia, by highlighting outstanding women in different research areas. This is particularly important in light of the ongoing poor gender distribution among invited speakers for many key conferences, as was for instance highlight in the recent controversy with regard to the speakers list for the 15th International Congress in Quantum Chemistry (ICQC), which is the triannual flagship conference of the International Academy of Quantum Molecular Science48,58, to name just one example.\n\nAs discussed in this contribution, a major contribution to the low percentages of female science professors is the existence of a “Matilda effect” in science, that manifests itself from a very early career stage, and which women fall as easily prey to the exercising of as men do. Here, there have been significant advances in strategies to address implicit bias in the workplace, as well as in funding and promotion panels and peer review (a quick internet search on this topic will provide countless hits), and I would also strongly recommend taking an implicit association test such as that provided by Harvard University (https://implicit.harvard.edu/implicit/takeatest.html) to test your own implicit biases. Unfortunately, by the very nature of being “implicit” we all carry some level of bias, and self-awareness and self-correcting for our biases can go a very long way towards fighting the Matilda effect in science.\n\n\nConclusion\n\nIn this opinion article, I have discussed at length the role in which explicit and implicit bias, both in terms of external perceptions and personal perceptions of one’s competence and ability, can play as barriers to female progression in academia. As a tenured faculty member working in computational biology (which is a research area which still maintains lower participation of women), I put my academic success strongly down to the fact that from an early stage, I had very strong role models giving me support and encouragement, and believing in my ability to achieve this. I believe, therefore, it is extremely important to give back to other younger colleagues, to give them the same opportunities and support to succeed in a system where the odds are still stacked against female academics. To facilitate this, I actively recruit and mentor highly promising young women to my research team, and take great pleasure from watching their own career success in turn. Here, I do my best to pay particular care to the knowledge that in the Matilda effect, women are just as biased as men. Amelia Earhart once said, “Women must try to do things as men have tried. When they fail, their failure must but be a challenge to others”59. Tremendous contributions have been made by structured programs to increase the presentation of woman in senior academic positions, such as the NSF Advance program60 in the US, or the Athena Swan program in the UK61. Ultimately, however, academia is comprised of each and every one of us, and it is the choices we make that will determine the future representation of women in the academic world.",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nI am not an expert in gender studies, but rather a scientist, writing from my own personal experiences, interpretation of the literature, and through extensive discussion with numerous colleagues. On this front, I would like to in particular like to thank Anna Krylov (University of Southern California), Mikael Widersten (Uppsala University), Sherry Mowbray (Uppsala University) and Karin Stensjö (Uppsala University) for their valuable insights into and helpful discussions about this topic.\n\n\nReferences\n\nSneider A: The new suffrage history: Voting rights in international perspective. History Compass. 2010; 8(7): 692–703. Publisher Full Text\n\nGary CS: Bildung and gender in nineteenth-century Bourgeois Germany: A cultural studies analysis of texts by women writers. 2008. Reference Source\n\nGöteborgs Universitetsbibliotek: Årtalslistor. [20th May 2016]. Reference Source\n\nChiu CS: Women in the shadows: Mileva Einstein-Marié, Margarete Jeanne Trakl, Lise Meitner, Milena Jesenká and Margarete Schütte-Lihotzky. Borchardt TE editor. New York: Peter Lang Publishing Inc.; 2008. Reference Source\n\nMaddox B: The double helix and the 'wronged heroine'. Nature. 2003; 421(6921): 407–8. PubMed Abstract | Publisher Full Text\n\nEssinger J: Ada's algorithm: How Lord Byron's daughter Ada Lovelace launched the digital age. London: Gibson Square Books Ltd.; 2013. Reference Source\n\nNobel Foundation: Enhancing X-ray Vision. [20th May 2016]. Reference Source\n\nHolloway M: Profile: Jane Goodall - Gombe's famous primate. Sci Am. 1997; 277: 42–4. Publisher Full Text\n\nKeats J: Hollywood star whose invention paved the way for Wi-Fi. 2011; [20th May 2016]. Reference Source\n\nSchwab K, Samans R, Zahidi S, et al.: The global gender gap report 2015. 10th anniversary edition. World Economic forum, 2015; [3rd April 2015]. Reference Source\n\nEuropean Commission: She figures 2012: Gender in research and innovation. Statistics and indicators. 2012. Reference Source\n\nSeierstad C, Healy G: Women's equality in the Scandinavian academy: A distant dream? Work Employ Soc. 2012; 26(2): 296–313. Publisher Full Text\n\nRice C: Why women leave academia. 2012; [20th May 2016]. Reference Source\n\nSilander C, Haake U, Lindberg L: The different worlds of academia: A horizontal analysis of gender equality in Swedish higher education. High Educ. 2013; 66(2): 173–88. Publisher Full Text\n\nCeci SJ, Williams WM: The mathematics of sex: How biology and society conspire to limit talented women and girls. New York: Oxford University Press; 2009. Reference Source\n\nHemel DJ: Summers' comments on women and science draw ire. 2005; [20th May 2016]. Reference Source\n\nAlberts B, Kirschner MW, Tilghman S, et al.: Rescuing US biomedical research from its systemic flaws. Proc Natl Acad Sci U S A. 2014; 111(16): 5773–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlocker N, Drodzewski D: Career progress relative to opportunity: How many papers is a baby 'worth'? Environ Plann A. 2012; 44(6): 1271–7. Publisher Full Text\n\nEpifanio M, Troeger VE: How much do children really cost? Maternity benefits and career opportunities of women in academia. Department of Economics, University of Warwick, 2013. Reference Source\n\nRafnsdóttir GL, Heijstra TM: Balancing work-family life in academia: The power of time. Gend Work Organ. 2013; 20(3): 283–96. Publisher Full Text\n\nTrussell DE: Pinstripes and breast pumps: Navigating the tenure-motherhood-track. Leisure Sci. 2015; 37(2): 160–175. Publisher Full Text\n\nMason MA: In the ivory tower, men only. 2013; [20th May 2016]. Reference Source\n\nMiller CC: The motherhood penalty vs. the fatherhood bonus. 2014; [cited 20th May 2016]. Reference Source\n\nKittelstrom A: The academic-motherhood handicap. 2010; [20th May 2016]. Reference Source\n\nWolfinger NH: For female scientists, there's no good time to have children. 2013; [20th May 2016]. Reference Source\n\nShen H: Inequality quantified: Mind the gender gap. Nature. 2013; 495(7439): 22–4. PubMed Abstract | Publisher Full Text\n\nFörsäkringskassan: Föräldrapenning. [2nd June 2016]. Reference Source\n\nEuropean Platform for Investing in Children: Denmark: Combining work and family life successfully. [2nd June 2016]. Reference Source\n\nGovernment of the Netherlands: Q&A pregnancy and maternity leave. [20th May 2016]. Reference Source\n\nFlüchtlinge BfMu: Elterngeld und Elternzeit. [20th May 2016]. Reference Source\n\nAngloinfo: Maternity leave and allowances. [20th May 2016]. Reference Source\n\nThe Wall Street Journal: Why Mom's time is different from Dad's time. [cited 20th May 2016]. Reference Source\n\nHansen AA, Cedstrand S, Gisselman M, et al.: Ojämställd arbetsborda: Föräldraledighetens betydelse för fördelning av betalt och obetalt arbete. Försäkringskassan; 2013. Reference Source\n\nBennhold K: In Sweden, men can have it all. 2010; [20th May 2016]. Reference Source\n\nAlbrecht J, Skogman Thoursie P, Vroman S: Glastaket och föräldrarförsäkringen i Sverige. Institutet för arbetsmarknads- och utbildningspolitisk utvärdering (IFAU); 2015. Reference Source\n\nHymowitz K: Longer maternity leave not so great for women after all. 2013; [20th May 2016]. Reference Source\n\nMerton RK: The Matthew Effect in Science: The reward and communication systems of science are considered. Science. 1968; 159(3810): 56–63. PubMed Abstract | Publisher Full Text\n\nRossiter MW: The Matthew Matilda effect in science. Soc Stud Sci. 1993; 23(2): 325–41. Publisher Full Text\n\nSteinke J, Lapinski MK, Crocker N, et al.: Assessing media influences on middle school-aged children's perceptions of women in science using the draw-a-scientist test (DAST). Sci Commun. 2007; 29(1): 35–64. Publisher Full Text\n\nBug A: Swimming against the unseen tide. Phys World. 2010; 23: 16–7. Publisher Full Text\n\nMoss-Racusin CA, Dovidio JF, Brescoll VL, et al.: Science faculty's subtle gender biases favor male students. Proc Natl Acad Sci U S A. 2012; 109(41): 16474–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWenneras C, Wold A: Nepotism and sexism in peer-review. Nature. 1997; 387(6631): 341–3. PubMed Abstract | Publisher Full Text\n\nVetenskapsrådet: Kvinnor på frammarsch i Vetenskapsrådets bidragsbeslut. 2014; [20th May 2016]. Reference Source\n\nAhlqvist V, Andersson J, Söderqvist L, et al.: En jämställd process? Vetenskapsrådet; 2015. Reference Source\n\nBornmann L, Mutz R, Daniel HD: Gender differences in grant peer review: A meta-analysis. J Informetr. 2007; 1(3): 226–38. Publisher Full Text\n\nDay TE: The big consequences of small biases: A simulation of peer review. Research Policy. 2015; 44(6): 1266–1270. Publisher Full Text\n\nSchroeder J, Dugdale HL, Radersma R, et al.: Fewer invited talks by women in evolutionary biology symposia. J Evol Biol. 2013; 26(9): 2063–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrylov A, Carter E, Gagliardi L: Stop gender discrimination in science. 2014; [20th May 2016]. Reference Source\n\nLarivière V, Ni C, Gingras Y, et al.: Bibliometrics: global gender disparities in science. Nature. 2013; 504(7479): 211–3. PubMed Abstract | Publisher Full Text\n\nCameron EZ, Gray ME, White AM: Is publication rate an equal opportunity metric? Trends Ecol Evol. 2013; 28(1): 7–8. PubMed Abstract | Publisher Full Text\n\nvan der Brink M: Scouting for talent: appointment practices of women professors in academic medicine. Soc Sci Med. 2011; 72(12): 2033–40. PubMed Abstract | Publisher Full Text\n\nvan der Brink M, Benschop Y: Gender practices in the construction of academic excellence: Sheep with five legs. Organization. 2011; 19(4): 507–24. Publisher Full Text\n\nScheltzer JM, Smith JC: Elite male faculty in the life sciences employ fewer women. Proc Natl Acad Sci U S A. 2014; 111(28): 10107–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVázquez-Cupeiro S, Elston MA: Gender and academic career trajectories in Spain: From gendered passion to consecration in a Sistema Endogámico? Employee Relat. 2006; 28(6): 588–603. Publisher Full Text\n\nHealey G, Ozbilgin M, Aliefendioglu H: Academic employment and gender: A Turkish challenge to vertical sex segregation. Eur J Ind Relat. 2005; 11(2): 247–64. Publisher Full Text\n\nYoung Academy of Europe, Uppsala SciLifeLab: Pathways to excellence: Experiences of illuminating women in science. 2015; [20th May 2016]. Reference Source\n\nKrylov AI: Women in theoretical/computational chemistry, material science, and biochemistry. [20th May 2016]. Reference Source\n\nInternational Academy of Quantum Molecular Science: The 15th ICQC: International Congress of Quantum Chemistry. [20th May 2016]. Reference Source\n\nEarhart A: Amelia Earhart: The official website. [20th May 2016]. Reference Source\n\nNational Science Foundation: ADVANCE: Increasing the participation and advancement of women in academic science and engineering careers. [22nd May 2016]. Reference Source\n\nEquality Challenge Unit: Athena Swan Charter. [22nd May 2016]. Reference Source"
}
|
[
{
"id": "14197",
"date": "07 Jun 2016",
"name": "Pernilla Wittung-Stafshede",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an excellent text on gender bias in academia with a focus on Sweden but with a general perspective. The topic is highly important, especially since people in many countries believe Sweden is a gender equal country.\nAs the author clearly demonstrates with facts and numbers, this is not true. The text is well written: it contains a historical perspective and the most famous studies demonstrating gender bias in academia are described.\nThe solution that senior women must mentor younger women and help them believe in themselves and pursue careers in academia is very good. Nonetheless, I believe more must be done at Swedish universities to achieve a true change in a foreseeable future. All faculty and administrators need to be educated about gender bias, both conscious and unconscious, and every academic leader must strive for equal treatment in every situation and decision. See below where I outlined some actions to take around the same topic: http://www.stemwomen.net/is-the-gender-gap-solved-in-liberal-sweden/",
"responses": [
{
"c_id": "2015",
"date": "07 Jun 2016",
"name": "Lynn Kamerlin",
"role": "Author Response F1000Research Advisory Board Member",
"response": "I would like to thank the reviewer for the positive assessment of my perspective piece, and also for the link to their related piece in STEM Women. It is clear from that piece that the problems I describe in my work only increase as one moves up the seniority ladder, and therefore it is refreshing to see the constructive 6-point plan at the end of the STEM Women piece. In particular, I fully agree with the importance of gender bias training, since the nature of implicit biases is that they are implicit, and therefore the people who carry them are not even aware that this is the case. I hope that debate about this issue increases, as well as proactive action, so we can move towards true equality of opportunity in academia."
}
]
},
{
"id": "14351",
"date": "14 Jun 2016",
"name": "Anna I. Krylov",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an excellent paper discussing the factors responsible for slow advancement of women in academe. The author points out to societal pressures and expectations, uneven distribution of family responsibilities, and unconscious biases. Her points are well documented and supported by data and numerous studies.\nAs far as biases are concerned, I would also mention Steinpreis et. al, Sex Roles, v. 41 (1999). In this study, identical academic resumes (differing only by the gender of the candidate) were mailed for expert evaluation for hireability and tenurability. The outcome was that the male name on the resume resulted in higher competence scores. Particularly disconcerting was that female tenure candidates were four times as likely to receive cautionary comments such as \"We would have to see her job talk\", \"I would need to see evidence that she had gotten these grants and publications on her own\", \"It is impossible to make such a judgment without teaching evaluations\".\nWhile the existence of biases has been well documented and is now reluctantly recognized in the community, the implications of societal pressures and cultural expectations are usually not openly discussed. I applaud the author for pointing out that what some consider to be family friendly practices, such as unnecessary long maternity leaves practiced in Nordic countries, are hugely detrimental for women's advancement and gender equality. What is needed is not long maternity leaves, but the availability of high-quality affordable child care and domestic help.\nIn my opinion, outdated societal expectations and cultural reality are primarily responsible for women dropping out from the workforce. Even when child-care options are available, there is an expectation that mothers have to provide personal day-to-day care to their children, be involved in school activities through volunteering, provide support for extra-curriculum activities, etc. These expectations stem from the culture of stay-home moms or women choosing mommy-track (which exists both in the US and in Europe) leads to over-parenting practices and creates a pressure for normal women to follow the suit. I believe we cannot achieve true gender equality as long as it is considered to be acceptable to be married to a stay-home wife.\nBesides the child-care practices, broader implications of this is that men in traditional marriages contribute to perpetuation of biases and create an unhealthy workplace environment. This is documented in the following study: Desai, Sreedhari D. and Chugh, Dolly and Brief, Arthur, The Organizational Implications of a Traditional Marriage: Can a Domestic Traditionalist by Night be an Organizational Egalitarian by Day? (March 12, 2012). UNC Kenan-Flagler Research Paper No. 2013-19. The quote below illustrates the point.\n\"Based on five studies with a total of 993 married, heterosexual male participants, we found that marriage structure has important implications for attitudes, beliefs, and behaviors related to gender among heterosexual married men in the workplace. Specifically, men in traditional marriage (married to women who are not employed) disfavor women in the workplace and are more likely than the average of all married men to make decisions that prevent the advancement of qualified women. Results show that employed men in traditional marriages tend to (a) view the presence of women in the workplace unfavorably, (b) perceive that organizations with higher numbers of female employees are operating less smoothly, (c) perceive organizations with female leaders as relatively unattractive, and (d) deny qualified female employees opportunities for promotions more frequently than do other married male employees.\"\nIn sum, this is an excellent viewpoint paper. I will recommend it to others as a valuable resource. My only criticism of the paper is the choice of the venue for its publication. The author should have submitted this paper to a reputable journal adhering to established publication and peer-reviewing practices.",
"responses": [
{
"c_id": "2029",
"date": "14 Jun 2016",
"name": "Lynn Kamerlin",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Thank you very much for the thoughtful and valuable comments on the paper, and also for the useful additional references. I am very glad that the reviewer appreciated the work, and also agree fully with all the additional points raised in the report. From the two referee reports it's clear that this is an issue that is significantly under-discussed, but that resonates strongly with other female professors as well, and I hope this encourages greater debate around (and constructive solutions to) this issue."
}
]
},
{
"id": "14201",
"date": "20 Jun 2016",
"name": "Sarah de Rijcke",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an excellent, well-written and well-researched opinion paper about barriers to female progression on the academic career ladder. Though Kamerlin carefully positions herself in the context of the Swedish research system and in the natural sciences, her observations (unfortunately) apply to many more contexts and fields.\nI would like to pick up on one point that Kamerlin discusses at quite some length: the issue of implicit bias. This is an issue facing many women from very early on in their academic career, as a result of deeply entrenched stereotypes about what constitutes a proper scientist, what a scientist looks like and what a scientist does. Kamerlin rightly points out how these implicit biases affect women's own career choices, in addition to the hiring and promotion procedures they are subjected to, and even the peer review of their grant proposals and papers.\nThere is one point about implicit bias that the author leaves untouched though in an otherwise broad-ranging and very comprehensive article. And this is the point about how academic merit itself is not objective and a-political, but heavily masculinised. I recommend reading the excellent article by Margaret Thornton on the contemporary re-masculinisation of the academy (Thornton, 2013), in which she shows how academic capitalism also exercises incidental gender effects. In today's highly competitive and precarious climate, Thornton argues, researchers of any gender are incentivised to focus mainly on masculinised performance measures that promote productivity and accumulation of capital. In the process, other - feminised - activities are seen as unproductive. This includes pastoral care for students, thinking, reading, anonymous reviewing, and mentoring of junior researchers. What counts is that which can be counted. Anything else is rendered as 'waste'.\nI hope that discussions about barriers to female career progression will also spearhead into an open debate about the kind of science we currently hold in highest regard in the first place. Does 'the gendered sub-text of technopreneurialism' (Thornton, 2013: 134) override more complex notions of quality and academic virtues? Does 'benchmark masculinity' (ibid: 136) affect the kinds of knowledge we produce under such conditions?\nThat said, I wholeheartedly agree with Kamerlin that we need more mentors, and more role models. And she is definitely one of them. Many researchers should find inspiration in how Kamerlin uses the networks and platforms at her disposal to further the discussion about the gender gap in the academy, and how she is putting into practice a different set of leadership skills, in order to make a difference.",
"responses": [
{
"c_id": "2036",
"date": "20 Jun 2016",
"name": "Lynn Kamerlin",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Thank you for your careful reading of the manuscript, and for your kind words, both of which are much appreciated. Thank you also for the valuable reference, which covers a point that is often overlooked, and I think actually very important in terms of understanding the parameters shaping our experiences as researchers. One can only hope that such researcher trickles down to practitioners to cause a paradigm shift towards a more healthy working environment for the next generation of academics."
}
]
},
{
"id": "14198",
"date": "24 Jun 2016",
"name": "Ruth Müller",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI enjoyed reading this opinion piece. The author, a researcher from the natural sciences, engages with different quantitative and qualitative studies from gender studies and science & technology studies and relates the results on how gender influences career development and recognition to her own experiences. She clearly states that this is the position she is writing from and hence adequately situates her contribution.\n\nI have minor comments to improve this opinion piece:\nFirst, I am not sure what purpose the section that discusses Ceci & William's work serves within the opinion piece; the section seems disconnected from the rest of the piece; I think this section could be left out or rewritten. For a rewrite I want to note that Ceci & Williams are not scholars in the field of gender studies, but rather study sex differences (e.g. in relation to I.Q. or career choice).\nSecond, the author draws our attention very clearly to how gender discrimination is not only a question of parenthood but based on deeply rooted stereotypical assumptions about the interests and capabilities of the dichotomously constructed categories of women and men. One could add that reframing gender discrimination as mainly a problem of working out the \"motherhood question\" and enabling and supporting motherhood actively obscures other aspects of discrimination, which, as the author elaborates, are very potent factor, too.\nThird, I would like to draw attention to how in the Turkey example the reliance of academic professors on domestic help may be very favorable for their careers, at the same this constellation redistributes domestic labour mainly among women and along categories of class and ethnicity. I would invite the author to mention that.\nForth, with regard to the \"Raising awareness of implicit bias\" section, I would invite the author to discuss the question if decision makers (referees for grants, hiring committees... etc) should undergo gender bias awareness training before being granted the power to decide? This has been a controversial question and I am interested in the authors opinions.\nLast, as the author, much to my pleasure, is interested in concrete action to better the situation, I would like to draw attention to a study by Trix & Psenka from 2003 that describes gender bias in recommendation letters. I find this study instructive for checking you own language in describing and assessing those who trust you to recommend them for career relevant positions. http://das.sagepub.com/content/14/2/191.abstract",
"responses": [
{
"c_id": "2067",
"date": "14 Jul 2016",
"name": "Lynn Kamerlin",
"role": "Author Response F1000Research Advisory Board Member",
"response": "I am delighted that the reviewer enjoyed reading the opinion piece, and appreciative of the constructive suggestions for further improving it. Based on the reviewer’s suggestions, a revision has now been submitted in which the following modifications have been made to the manuscript: With regard to the work of Ceci & Williams, although the authors are not scholars in the field of gender studies, the views they summarize align fairly well with my “on the ground” experiences. This has now been stated explicitly in the manuscript to explain what that section is doing there. A statement has been added cautioning against allowing the “motherhood question”, which remains a major challenge, to take over and obfuscate other aspects of discrimination, which are also very potent, and thus why it’s important to approach this issue with care. A statement has been made about the problems with an over-reliance on domestic help simply redistributing the problem amongst women and other categories of class and ethnicity. A statement has been made about implicit bias training for decision makers before sitting on grant and hiring committees. Although this is controversial and only raises awareness of rather than solves problems, nevertheless it’s an important first step. Here, I would like to also point out that implicit bias is not only a gender issue, but involves also bias based on ethnicity, academic rank, geographic location, and any other number of variables, and poses a significant problem in grant evaluation and recruitment processes. Therefore, implicit bias training should not be limited to gender awareness alone. A short paragraph discussing gender bias in recommendation letters (based on the study of Trix and Penska from 2003) has been now included in the manuscript."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1224
|
https://f1000research.com/articles/5-282/v1
|
04 Mar 16
|
{
"type": "Research Note",
"title": "Acceptance of animal research in our science community",
"authors": [
"Konstantin Bergmeister",
"Bruno Podesser",
"Konstantin Bergmeister"
],
"abstract": "Animal research is debated highly controversial, as evident by the “Stop Vivi-section” initiative in 2015. Despite widespread protest to the initiative by researchers, no data is available on the European medical research community’s opinion towards animal research. In this single-center study, we investigated this question in a survey of students and staff members at the Medical University of Vienna. A total of 906 participants responded to the survey, of which 82.8% rated the relevance of animal research high and 62% would not accept a treatment without prior animals testing. Overall, animal research was considered important, but its communication to the public considered requiring improvement.",
"keywords": [
"Animal research",
"survey",
"acceptance animal research",
"Stop vivi-section"
],
"content": "Introduction\n\nAnimal research is still debated highly controversial and lately attracted great attention as over 1.1 million European citizens signed the “Stop Vivi-section” initiative in 2015, demanding the stop of all animal research1. Alarmed by the potential consequences opinion leaders made efforts to illustrate the need for animal experiments for medical progress2,3. However, does the European medical research community stand united behind animal research?\n\n\nMethods\n\nIn an internal survey at the Medical University of Vienna we investigated the positions towards animal research of 10335 (M.D. and Ph.D.) students and 3824 medical staff members. The survey was conducted using the MedCampus system (CAMPUSOnline, Graz, Austria) of the Medical University of Vienna, accessible to all students and staff members. The survey was conducted over a period of four weeks in November 2015. Statistical analyses were conducted using SPSS (V.21, IBM Corp, US).\n\nEthics committee approval: Approval was obtained from the Medical University of Vienna’s data privacy committee.\n\n\nResults\n\nA total of 906 participants responded to the survey, resulting in a response rate of 6.38%. Participants were 36.5% staff members and 63.5% students, of which 43% previously had personal experience with animal experiments. The relevance of animal models for research was rated high (8–10 on a scale 1–10; 1 being lowest) by 82.8%, and 62% would not accept a treatment without prior animals testing (Figure 1, left). These results were similar to a 2011 Nature poll4 with 980 participants and a 2014 survey by the American Association for the Advancement of Sciences5. In our cohort, participants rated the society’s acceptance of animal research low (4.24 ±1.77, scale 1–10; 1 being lowest) as well as the current communication to the public on medical advances derived from animal research (4.37 ±2.22, scale 1–10; 1 being lowest). Consequently, 75.4% believed the public should receive better information about the benefits, necessities and legislation of animal experiments (Figure 1, right).\n\nLeft: A majority of participants would not accept a treatment that has not been previously tested in animal models. Right: The need for better information about animal research for the public was rated high by 75% of the participants.\n\n\nDiscussion\n\nIn this study, we assessed the opinions of our faculty members and students towards animal research. Overall, our study population considered animal research important for medical progress. In addition, we see a clear mission to improve communication to the public about animal experiments. Moreover, scientists need to improve the communication of complex results into a language that is understood by society and colleagues alike. Limitations of this study were the small number of participants and being a single-center survey. A comparable nature study4 from 2011 had a relatively lower response rate (approximately 4.9%) and a similar total number of 980 participants.\n\nIn conclusion, this single-center study provides first survey results of students and medical faculty members towards animal research. Based on the interesting results, we plan to extend this study to other institutions and thereby provide an overview of the European medical community’s opinion towards animal research.\n\n\nData availability\n\nF1000Research: Dataset 1. Word file containing survey questions in original German language and translated to English, 10.5256/f1000research.8169.d1152196\n\nF1000Research: Dataset 2. Excel file containing anonymized responses to the survey, 10.5256/f1000research.8169.d1152207",
"appendix": "Author contributions\n\n\n\nKB and BP both designed the study, and collected, analysed and interpreted the data. KB carried out the literature search and wrote the manuscript and prepared the figures, while BP revised the manuscript critically.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Christian Doppler Research Association.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBalls M: The European Citizens’ Stop Vivisection Initiative. Altern Lab Anim. 2015; 43(3): 147–50. PubMed Abstract\n\nMohdin A: The Stop Vivisection Initiative – Trying to Ban European Animal Research. 2015. Reference Source\n\nEARA: Statement supporting European Directive 2010/63/EU (“Directive”) on the protection of animals used for scientific purposes. 2016. Reference Source\n\nCressey D: Animal research: Battle scars. Nature. 2011; 470(7335): 452–3. PubMed Abstract | Publisher Full Text\n\nFunk C, Rainie L, Page D: Public and Scientists’ Views on Science and Society. 2015. Reference Source\n\nBergmeister K, Podesser B: Dataset 1 in: Acceptance of animal research in our science community. F1000Research. 2016. Data Source\n\nBergmeister K, Podesser B: Dataset 2 in: Acceptance of animal research in our science community. F1000Research. 2016. Data Source"
}
|
[
{
"id": "13344",
"date": "26 Apr 2016",
"name": "Thomas Butts",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments This report communicates the details of a small, but not insignificant survey of doctoral students’ and professional scientists’ from the Medical University of Vienna attitudes towards animal research. The findings, perhaps not surprisingly, reveal strong support for animal research within the life science research community at this single institution, but importantly also highlight an awareness of the short-comings of communicating this necessity to the public in the light of recent political movements in opposition to research using animals. The data in reference to the use of a given medical treatment without previous testing on animals is particularly interesting and will contribute much to the debate. My sole concern with this research is the generality of the questions. For example, a distinction ought to be made in reference to the type and severity of treatment in the survey. Nevertheless, the data presented will contribute to public (and scientific) debate. The authors are correct in identifying the need for a more systematic survey amongst European life science institutions and professional groups that will inform this important area of public debate. Specific comments Introduction Line 1: ‘’…still debated highly controversial and lately…’’ should read ‘’…still debated, highly controversial, and lately has attracted…’’ Methods Paragraph 2, line 1: ‘’resulting’’ should be changed to ‘’representing’’ Line 10: ‘’the’’ should be deleted",
"responses": []
},
{
"id": "13710",
"date": "11 May 2016",
"name": "David B. Lumenta",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSurvey among doctoral students and staff members (1 university) on their own and general public's perception of animal experiments with 6.38% response rate. The presented questionnaire was general without requiring too much detail from respondents, which I found sufficient.The results are similar to previous research (as cited by the authors), the used methodology sound, and the conclusions, notably the need for a more systematic review among institutions in Europe on this topic, well balanced.",
"responses": []
},
{
"id": "14281",
"date": "10 Jun 2016",
"name": "David Bernhard",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study is well performed and of high interest for the scientific community. Similar studies in the general population would be very valuable and could help in the discussion and communication of the need for animal experiments.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-282
|
https://f1000research.com/articles/5-251/v1
|
01 Mar 16
|
{
"type": "Research Article",
"title": "Mast cell activation is enhanced by Tim1:Tim4 interaction but not by Tim-1 antibodies",
"authors": [
"Binh Phong",
"Lawrence P. Kane",
"Binh Phong"
],
"abstract": "Polymorphisms in the T cell (or transmembrane) immunoglobulin and mucin domain 1 (TIM-1) gene, particularly in the mucin domain, have been associated with atopy and allergic diseases in mice and human. Genetic- and antibody-mediated studies revealed that Tim-1 functions as a positive regulator of Th2 responses, while certain antibodies to Tim-1 can exacerbate or reduce allergic lung inflammation. Tim-1 can also positively regulate the function of B cells, NKT cells, dendritic cells and mast cells. However, the precise molecular mechanisms by which Tim-1 modulates immune cell function are currently unknown. In this study, we have focused on defining Tim-1-mediated signaling pathways that enhance mast cell activation through the high affinity IgE receptor (FceRI). Using a Tim-1 mouse model lacking the mucin domain (Tim-1Dmucin), we show for the first time that the polymorphic Tim-1 mucin region is dispensable for normal mast cell activation. We further show that Tim-4 cross-linking of Tim-1 enhances select signaling pathways downstream of FceRI in mast cells, including mTOR-dependent signaling, leading to increased cytokine production but without affecting degranulation.",
"keywords": [
"Tim-1",
"Tim-4",
"mast cells",
"phosphorylation",
"ribosomal S6"
],
"content": "Introduction\n\nT-cell immunoglobulin and mucin domain (TIM)-1 belongs to a family of type I transmembrane proteins with essential roles in immune regulation. Tim-1 has an N-terminal IgV domain, a mucin-like domain with multiple potential sites for O-linked glycosylation, followed by a stalk domain with potential N-linked glycosylation sites, a transmembrane domain, and a short cytoplasmic tail. Tim-1 was initially identified as a cellular receptor for hepatitis A virus (HAVCR1)1. The TIM-1 gene was positionally cloned from a locus on the same chromosome but distinct from the IL-4 gene cluster that is commonly associated with Th2-biased immune responses2. Polymorphisms in murine Tim-1 are associated with susceptibility to airway hyper-reactivity (AHR), a hallmark of asthma, and increased Th2 cytokine production2. Similarly, polymorphisms in human TIM-1 have been linked to atopic diseases including asthma, allergic rhinitis and atopic dermatitis3. The atopy connection is particularly intriguing since genetic variations in human TIM-1 modify susceptibility or resistance to allergy, but only in individuals sero-positive for HAV4. These findings suggested that Tim-1 has a role in regulation of immune responses to atopic diseases.\n\nMechanistically, Tim-1 was shown to co-stimulate T effector cell proliferation, with preferential effects on Th2 cytokine production. Thus, the high affinity agonistic monoclonal antibody (mAb) 3B3 was reported to inhibit the induction of respiratory tolerance in an AHR model5, and to enhance both T cell proliferation and cytokine production in vitro and in vivo5,6. Antibodies recognizing distinct epitopes of Tim-1 either enhanced or attenuated lung inflammation7. We further showed that ectopic expression of Tim-1 in T cells stimulated in vitro under “neutral” conditions promoted generation of more IL-4- rather than INF-γ production6. Tim-1 can also enhance NFAT/AP-1-dependent transcription in T cells activated by TCR crosslinking, suggesting that Tim-1 functions as a co-stimulatory molecule for T cell activation8. Similarly, co-stimulatory function of Tim-1 was also observed after interaction of Tim-1 with its ligand Tim-4, which is primarily expressed on APCs9,10.\n\nRegarding signaling pathways coupled to Tim-1, we showed that tyrosine 276 in the cytoplasmic tail of Tim-1 could be phosphorylated in an Lck-dependent manner. This allows for recruitment of the p85α and β subunits of the PI3K, leading to activation of the downstream kinase Akt and subsequent activation of the transcription factors NFAT and AP-18. Administration of the agonistic Tim-1 antibody 3B3 induces expression of early activation markers CD69 and CD25 as well as IL-2 production8. Other groups have demonstrated that ligation of Tim-1 by Tim-4 can activate the ERK/MAPK pathway and enhance T cell survival by up-regulating the anti-apoptotic protein BcL-xL9. Additional studies revealed that Tim-1 could co-cap with CD3 on human T cells11. Tim-1 ligation on T cells has also been reported to induce tyrosine phosphorylation of the linker for activation of T cells (LAT) and the TCR-proximal Syk family tyrosine kinase Zap709. Taken together, these findings suggested that Tim-1 may interact with proximal TCR signaling complexes.\n\nIn addition to T cells, Tim-1 also has regulatory functions on other non-immune and immune cell types. Tim-1, also known as kidney injury molecule (KIM)-1, is upregulated on renal proximal tubules and shed upon acute renal failure12. Apoptotic cell recognition by Tim-1, specifically on natural killer T (NKT) cells, may induce AHR in response to respiratory syncytial virus- or ozone-induced experimental asthma13. Tim-1 has recently been shown to be expressed by IL10-secreting regulatory B cells and Tim-1 signaling is required for the induction and maintenance of these cells14,15. Specifically, the Tim-1 mucin domain is required for IL-10 production in response to phosphatidylserine (PS) binding and allograft tolerance14,15. Tim-1 is also constitutively expressed on bone marrow-derived (BMMC) and peritoneal (PMC) mast cells. Cross-linking of Tim-1 by Tim-4 enhanced IgE plus antigen-stimulated (IgE/Ag) production of Th2 type cytokines16. However, the mechanisms by which Tim-1 modulates mast cell functional responses are currently unknown.\n\nMast cells are among the first responders of immune responses against pathogens and allergens. They have the capacity to secrete a multitude of pro- and anti-inflammatory factors that regulate allergic inflammation, pathogen defense, and anti-tumor immunity17. Given the genetic and functional connection of Tim-1 to allergy and hypersensitivity and the sentinel role of mast cells in atopy, it is important to determine how Tim-1 signaling contributes to the high affinity Fc receptor for IgE (FcεRI)-mediated mast cell activation. Here we demonstrate that Tim-1 promotes NF-κB and NFAT/AP1 transcriptional activation, leading to enhanced IL-6 promoter activation and cytokine production in IgE/Ag-stimulated mast cells. Using BMMCs generated from a mouse strain lacking the Tim-1 mucin domain (Tim1Δmucin), we show that this co-stimulatory effect is independent of the Tim-1 mucin domain. Finally, we show that Tim-1, in contrast with Tim-3, acts more distal to FcεRI to enhance S6 activation, without affecting proximal FcεRI signaling. Overall, our findings provide a mechanistic explanation for the co-stimulatory effects of Tim-1 signaling on FcεRI-mediated mast cell activation.\n\n\nMethods\n\nMonoclonal anti-dinitrophenyl (DNP), IgE isotype, clone SPE-7 (Cat No. D8406), DNP32-HSA (Cat No. A6661), anti-FLAG M2 antibody (Cat No. F1804), cyclosporine A (Cat No. C1832), and 4-Nitrophenyl N-acetyl-β-D-glucosaminide (pNAG) (Cat No. N9376) were purchased from Sigma-Aldrich (St. Louis, MO). DNP5-BSA (Cat No. D5050) was from Biosearch Technologies (Petaluma, CA). Monoclonal antibodies to murine Tim-1 (3B3, RMT1-10, 5G5, 5F12) and purified Tim4-Fc were obtained from Vijay Kuchroo (Harvard Medical School). Human IgG as isotype control for Tim4-Fc was purchased from Jackson ImmunoResearch Laboratories (Cat No. 009-000-003, West Grove, PA). Monoclonal antibodies to BALB/c Tim-1 (1H9, 3A2, 4A2, 4B2, 4G8, 5D1) were originally developed at Biogen and were obtained under an MTA with CoStim Pharmaceuticals (Cambridge, MA). Purified rat IgG2a as isotype control for Tim-1 antibodies was purchased from eBioscience (Cat No. 14-4321-85, San Diego, CA). Phospho-specific antibody to ERK (T202/Y204) was from BD Biosciences (20 μl per test), (Cat No. 612566, San Jose, CA). Phospho-specific antibodies to Syk (Y519/520) (1:400, Cat No. 2710), and S6 (S235/236), (1:150, Cat No. 4851) were obtained from Cell Signaling Technology (Danvers, MA). Anti-mouse Fc block (2.4G2) was purchased from BD Biosciences (Cat No. 553141, San Jose, CA). IL-6 luciferase reporter constructs were obtained from Sarah Gaffen (University of Pittsburgh), originally from Oliver Eickelberg (Helmholtz Zentrum Munchen).\n\nAll studies were performed in accordance with University of Pittsburgh Institutional Animal Care and Use Committee procedures. Specifically, mice were housed four males or five females per polycarbonate cage in a 12-hour light/dark cycle. Food and water were provided ad libitum. Cages and bedding were changed every seven days. Mutant mice lacking Tim-1 mucin domain (Tim-1Δmucin) were obtained from David Rothstein (University of Pittsburgh), and were originally from Vijay Kuchroo (Harvard Medical School). Age-and sex-matched wild-type C57BL/6 were purchased from the Jackson Laboratory (Bar Harbor, ME) as control.\n\nBone marrow cells from C57BL/6 Tim-1 wild-type (WT) and mutant (Tim-1Δmucin) were generated as described previously18. MC/9 mast cells were cultured in DMEM supplemented with 10% BGS, 2-ME, Pen/Strep with Glutamine, and 10% IL3-conditioned media.\n\nBMMCs (1 × 105 or 1 × 106 cells for cytokine measurement and 3 × 105 cells for phospho-flow) were sensitized overnight with 1 μg/ml IgE without IL-3 conditioned media. Cells were stimulated with DNP32-HSA or DNP5-BSA. Six or twenty-four hours post stimulation, supernatants were collected and assayed for murine IL-6 and TNF-α by ELISA (BioLegend). For phospho-flow staining, stimulated cells were fixed in 1.5% paraformaldehyde for ten minutes at room temperature and permeablized with ice cold methanol for thirty minutes. Incubation with phospho-specific antibodies were performed per manufacturer’s instructions. Flow acquisition was performed on Fortessa or LSRII (BD Biosciences), and data were analyzed using FlowJo software version 8.7 (Tree Star).\n\nBMMCs (2.5 × 105 cells) were sensitized and subsequently stimulated for thirty minutes in Tyrode’s buffer (135mM NaCl, 5mM KCl, 5.6mM glucose, 1.8mM CaCl2, 1mM MgCl2, 20mM HEPES, and 0.5mg/ml BSA). Supernatants (stimulated release) were collected and cells were lysed (content) with 0.5% Triton-X100 in PBS for fifteen minutes on ice. Content and stimulated release fractions were incubated with 1mM pNAG substrate for 1 hour at 37°C. Carbonate buffer (0.1M, pH 9.0) was added to stop reaction and absorbance was obtained at 405nm on a plate reader (BioTek ELx808). Percentage of beta-hexosaminidase release was calculated as (releasestimulated/contenttotal) × 100.\n\nDegranulation was measured as described previously18,38. Briefly, BMMCs were loaded with 0.1μM of Lysotracker Deep Red (Invitrogen) for thirty minutes at 37°C prior to sensitization with 0.5 μg/ml IgE for 1 hour. Ninety minutes after antigen cross-linking by DNP32-HSA, cells were collected and stained with Annexin V (BioLegend). %degranulation was determined as percentage of BMMCs that is AnnexinV+Lysotrackerlo. Flow acquisition was performed on Fortessa or LSRII (BD Biosciences), and data were analyzed using FlowJo software version 8.7 (Tree Star).\n\nMC/9 mast cells (15 × 106) were transfected with 15 μg of IL6-luc, NF-κB-luc, or NFAT/AP1-luc with 5 μg of pCDEF3 (empty vector), FLAG-tagged Tim-1 full length (FL), Tim-1 cytoplasmic deletion (Δcyto), or Tim-1 tyrosine mutant (Y276F). Electroporation was performed at 290V, 950 μF using a Gene Pulser II apparatus (Bio-Rad). Cells were collected twenty-four hours post-transfection and stimulated with 0.5 μg/ml IgE and 30 ng/ml or 100 ng/ml of DNP32-HSA as antigen for six hours. Luciferase assays were performed as previously described39.\n\nAll statistical analyses was performed using Prism version 6.0 (GraphPad Software). Paired, unpaired, two-tailed Student’s t-test, one-way and two-way ANOVA with multiple comparison were used for data analysis and calculation of p values, as appropriate.\n\n\nResults\n\nC57BL/6 BMMCs was incubated with 0.5 μg of Tim4-Fc or human IgG (iso) on ice for 20 minutes followed by flow cytometry analysis (A). BMMCs were sensitized overnight with 1 μg/ml of IgE and stimulated with 50 ng/ml of DNP32-HSA, alone or with 5 μg/ml of Tim4-Fc or isotype control Fc (“iso”) for six hours prior to IL-6 measurement by ELISA (B). BMMCs were sensitized with IgE and stimulated with DNP32-HSA in the presence of increasing amount of Tim4-Fc (0.1–50 μg/ml) for thirty minutes prior to degranulation measurement by means of beta-hexosaminidase release (C). Cells were loaded with Lysotracker Deep Red, sensitized with IgE, and stimulated with antigen plus isotype control or Tim4-Fc for ninety minutes prior to Annexin V staining and flow cytometry analysis (D). BMMCs were sensitized with IgE and stimulated with DNP32-HSA (E) or DNP5-HSA (F) alone or in conjunction with indicated amount of control Fc (“iso”) or Tim4-Fc for six hours. Anti-mouse Fc blocking Ab (2.4G2) was added as indicated for ten minutes prior to antigen and antibody stimulation. IL-6 cytokine production was determined by ELISA. Results are representative of three independent experiments with duplicates in each (B, E, F), average of four experiments (C), and at least three experiments (D). *p<0.05.\n\nMast cells constitutively express surface Tim-1 and Tim-3, but not Tim-2 or Tim-416. Tim-1 cross-linking by Tim4-Fc was shown to enhance cytokine production in a dose-dependent manner without affecting degranulation16. We observed similar effects of Tim4-mediated IL-6 production in IgE/Ag-stimulated BMMCs (Figure 1B). In addition, we quantified the degranulation response by measuring beta-hexosaminidase release as well as with a flow cytometry-based assay that tracks both Annexin V binding to exposed PS and loss of Lysotracker staining, due to granule release (AnnexinV+Lysotrackerlo). The latter method has the advantage of a high signal-to-noise ratio, and as such has been a robust assay for evaluating IgE/Ag-stimulated mast cell degranulation18. Thus, we showed that varying the concentration of Tim-4, along with suboptimal antigen concentration to observe potential co-stimulation, did not have an effect on the immediate degranulation response (Figure 1C–D). These results suggest that Tim-1 ligation by Tim-4 can exert differential effects on the immediate degranulation response, vs. late-phase cytokine production.\n\nAntigen valency and concentration have been shown to control the outcomes of FcεRI engagement in not just quantitative, but also qualitative, fashions19. Specifically, low antigen concentration or valency will activate only positive FcεRI signaling pathways, while high antigen valency or concentration will preferentially engage negative signaling components downstream of FcεRI. This is due to activity of the Src family kinase Lyn as a positive and/or negative regulator of FcεRI signaling at low or high antigen valency, respectively19. We stimulated BMMCs with low (DNP5-BSA) or high (DNP32-HSA) potency antigens in the presence of Tim4-Fc or isotype control and assessed whether Tim-1 ligation could contribute to receptor signaling intensity. Thus, at high antigen valency, which also activates negative feedback of antigen receptor signaling, Tim-4 was able to maintain high IL-6 production but not at low antigen valency that induces robust antigen receptor signaling (Figure 1E–F). Specifically, increasing the amount of Tim-4 further promoted cytokine secretion under high valency antigen stimulation, i.e. under conditions where the negative signaling loop is triggered. To exclude the possibility that Fcγ receptor binding may interfere with Tim1-Tim4 interaction, we compared IL-6 release by BMMCs, with or without addition of Fc blocking antibody, and found no differences, at least at the concentration of Tim4-Fc used throughout this study (5 μg/ml). These findings indicate that Tim-1 ligation can modulate the intensity of the antigen-induced positive FcεRI signaling pathways and may be able to bypass the negative feedback signaling loop controlled by Lyn.\n\nSeveral antibodies against Tim-1 have been tested and showed no effects on either IgE/Ag-induced mast cell degranulation or cytokine production16. Of significance are the mAbs 3B3 and RMT1-10, which have been termed “agonistic” and “antagonistic” antibodies, respectively, due to their ability to enhance or inhibit effector T cell activation20. Similarly, we did not observe any dose-dependent effects of 3B3, RMT1-10 or the IgV domain-binding mAb 5F12 on levels of IL-6 and TNF-α secreted by IgE-sensitized and Ag-stimulated BMMCs (Figure 2A–D). We examined several other antibodies, generated against the BALB/c allele of Tim-1, which were shown to either exacerbate or ameliorate Th2-dependent OVA-induced lung inflammation in mice7. Specifically, mAbs 1H8 and 3A2 bind to distinct epitopes near an N-linked glycosylation site in the stalk region and have agonistic and antagonistic activity, respectively. The mAb 4A2 binds to the IgV domain of Tim-1 and reduces lung inflammation and pathology7. However, aside from 4G8, we did not detect significant binding of these antibodies to BALB/c Tim-1. While 4G8 recognized surface Tim-1, it did not alter Ag-mediated cytokine production in mast cells (Figure 2E). Even though these antibodies were raised against the BALB/c allele of Tim-1, we also observed binding of 4G8 to C57BL/6 Tim-1 (Figure 2F). There was a small, although statistically insignificant, increase in IL-6 production when 4G8 was used in co-stimulation with FcεRI crosslinking by IgE/Ag (Figure 2F). These results further support the concept that the effects of antibody modulation of Tim-1 are cell-type and context-dependent7,16.\n\nBL/6 or BALB/c BMMCs (1 × 106) as indicated were sensitized with IgE overnight and stimulated with DNP32-HSA in the presence of isotype control or monoclonal antibodies against Tim-1 3B3 (A–B), RMT1-10 (C–D), 3A2, 4A2, 4B2, 4G8, 1H9 (E) for six hours. The indicated antibodies were incubated with BALB/c BMMCs for 30 minutes on ice followed by anti-rat IgG-Alexa-647 secondary antibody, prior to flow cytometry analysis (E). Culture supernatants were analyzed for IL-6 and TNF-α by ELISA. BL/6 BMMCs were incubated with 4G8, 4A2 and 5D1 antibodies followed by anti-rat IgG-Alexa 647 secondary antibody prior to flow cytometry analysis (F). BL/6 BMMCs sensitized with IgE overnight and stimulated with DNP32-HSA together with either isotype control or the indicated antibodies for six hours prior to IL-6 measurement by ELISA (F). Results are representative of three (A–D) and two (E–F) independent experiments performed in duplicates.\n\nWe previously demonstrated that transient expression of Tim-1 co-stimulated TCR/CD28-mediated transcriptional activation of IL-4 and IFN-γ production and NF-AT/AP1-dependent transcription. This co-stimulatory activity was dependent on tyrosine 276 in the Tim-1 cytoplasmic tail6. Similarly, ectopic expression of Tim-1 on MC/9 mast cells was able to enhance IgE/Ag-stimulated NF-κB transcriptional activation (Figure 3A). This enhancement was abrogated when the cytoplasmic tail of Tim-1 was deleted or when a tyrosine-phenylalanine mutant (Y276F) was used (Figure 3A). Polymorphisms in Tim-1 have been associated with differential responses to OVA-induced allergic asthma2. We found that both isoforms of Tim-1 (BL/6 and BALB/c) could significantly enhance activation of an NF-κB transcriptional reporter to a comparable extent (Figure 3B). Consistent with findings in T cells and effects on mast cell cytokine production, transient expression of Tim-1 also up-regulated NF-AT/AP1 and IL-6 promoter activation, in a tyrosine phosphorylation-dependent manner (Figure 3C–D). Furthermore, using IL-6 promoter deletion constructs, we showed that Tim-1 mediated enhancement of transcriptional activation and subsequent production of IL-6 through activation of NF-κB and AP-1 transcription factors (Figure 3E). Unlike Tim-4 crosslinking of Tim-1, addition of anti-FLAG antibody could not further promote reporter activity, suggesting either that Tim-1 may have other ligands on MC/9 mast cells or that Tim-1 can homo-dimerize after ectopic expression, via its heavily glycosylated mucin domain, leading to downstream signaling.\n\nMC/9 mouse mast cells were transfected with empty vector (pCDEF3), BL/6 or BALB/c Tim-1 (full length (FL)), BL/6 Tim-1 lacking the cytoplasmic region (Δcyto) or full length Tim-1 (BL/6) with tyrosine to phenylalanine mutated at tyrosine 276 (Y276). Transfected cells were stimulated with 0.5 μg/ml IgE plus either 30 ng/ml or 100 ng/ml DNP32-HSA, with or without addition of anti-FLAG antibody for six hours. MC/9 mast cells were co-transfected NF-κB (A–B), NFAT/AP1 (C), and IL-6 (D) luciferase reporters. MC/9 mast cells were transfected with full length BL/6 Tim-1 along with indicated IL-6 luciferase reporters and stimulated as described (E). Results are representative of three independent experiments performed in triplicate. *p<0.05, **p<0.005, ****p<0.00005.\n\nOur findings thus far suggest that the Tim1-Tim4 interaction augments FcεRI signaling itself, rather than acting through a parallel pathway, since Tim-4 treatment alone does not induce any detectable cytokine production or degranulation. While the Tim1-Tim4 interaction has been attributed primarily to the IgV domain, Tim-4 has also been proposed to bind to the Tim-1 mucin domain10. Intriguingly, the genetic linkage of Tim-1 to allergies and asthma is associated with polymorphisms in the mucin domain. Using a mutant mouse lacking only the Tim-1 mucin domain (Tim-1Δmucin)21, we determined whether the Tim-1 mucin domain is necessary to relay the co-stimulatory effects of Tim-4 binding. We noted no obvious defects in mast cell development or maturation, when BMMC were generated from Tim-1Δmucin bone marrow. Thus, WT and Tim-1Δmucin BMMCs expressed comparable levels of surface Tim-1, based on staining with IgV-binding Tim-1 antibodies (Figure 4A). Tim-1Δmucin BMMCs exhibited a similar extent of degranulation in response to antigen stimulation, compared to WT BMMCs, which was not altered by Tim-1 ligation (Figure 4B–C). Next, we examined the ability of Tim-1Δmucin BMMCs to secrete cytokines in response to IgE/Ag. Tim-1Δmucin mast cells appeared to respond normally to antigen stimulation, and also to Tim4-mediated co-stimulation (Figure 4D). Using different batches of BMMCs over the course of multiple experiments, we were not able to consistently observe any major difference between WT and Tim-1Δmucin BMMCs. We did observe some variation in the amount of IL-6 produced by different batches of BMMCs, but this appeared to be largely due to the relative maturation status of the cells (Figure 4E–H). These results indicated that the mucin domain of Tim-1 is not required for normal mast cell responses and also that Tim-4-mediated mast cell activation does not appear to involve the mucin domain of Tim-1, but rather its IgV domain.\n\nTim-1 surface expression and maturity of WT and Tim-1Δmucin BMMCs were determined by FcεRI and c-kit staining (A). WT and Tim-1Δmucin BMMCs were sensitized overnight with IgE and stimulated with 10 ng/ml, 50 ng/ml, 100 ng/ml DNP32-HSA, and 5 μg/ml Tim4-Fc. Mast cell degranulation was measured by Annexin V and Lysotracker staining (B–C). Results are average of three (B) and two (C) independent experiments. Culture supernatants were collected and analyzed for IL-6 by ELISA (D). Using six separate batches of BMMCs. WT and Tim-1Δmucin BMMCs were stimulated with indicated amount of antigen with either 5 μg/ml isotype control Fc (iso) or Tim4-Fc for six hours (E–H). Results are comparison between WT and Tim-1Δmucin BMMCs from six independent experiments using six separate batches of BMMCs. *p<0.05.\n\nNext, we assessed the potential signaling pathways utilized by Tim1-Tim4 interaction downstream of FcεRI signaling to upregulate mast cell cytokine production. Using a Nur77GFP reporter mouse, we previously showed that unlike the related family member Tim-3, engagement of Tim-1 did not enhance FcεRI signal intensity, thereby showing that Tim-1 cross-linking does not augment antigen receptor-proximal signaling18. We moved on to explore signaling pathways both proximal and distal to the FcεRI complex, for any effects of Tim-1. Thus, the Zap70-related kinase Syk is an FcεRIγ-associated activator integral to activation of LAT, SLP76, PLC-γ and other adaptor molecules essential for signal transduction downstream of FcεRI22. Phosphorylation of Syk was not further increased by Tim-4 (Figure 5A). Similarly, we observed robust phosphorylation of ERK and Akt upon IgE/Ag-induced activation that was not affected by Tim-1 engagement (Figure 5B and data not shown). These results demonstrate that MEK/ERK and PI3K/Akt pathways are not involved in enhancement of Ag-mediated mast cell function by Tim-1, even though we observed that phosphorylation of Tim-1 cytoplasmic tail by TCR activation led to recruitment of p85 subunit of PI3K8. Nevertheless, we did detect a significant increase in phosphorylation of ribosomal protein S6, an important target of the PI3K/mTOR pathway regulating cell growth, survival, metabolism, and protein synthesis in mast cells23. Mast cells exhibited robust phosphorylation of S6 (~80% of BMMCs) one hour post-Ag stimulation, which did not increase further with Tim-4 addition. However, Tim-4 treatment was able to maintain a significant percentage of pS6-positive BMMCs for as long as four hours, even as the Ag-triggered signal returned to basal levels (Figure 5C). Overall, these results support a positive role of Tim-1 activation by Tim-4 to sustain mTOR-dependent mast cell metabolism and protein synthesis, leading to enhanced cytokine production.\n\nBL/6 BMMCs were sensitized overnight with IgE and stimulated with DNP32-HSA in the presence of isotype control or Tim4-Fc for the indicated time. Syk phosphorylation (Y519/520) (A), pErk (T202/Y204) (B), and pS6 (S235/236) (C) were analyzed by phospho-flow cytometry. Results are average of two independent experiments (A–B) or three independent experiments (C). *p<0.05, ***p<0.0005.\n\n\nDiscussion\n\nMast cells constitutively express high levels of cell-surface Tim-1, a molecule with co-stimulatory effects on many immune cell types, but with unclear mechanisms of action. Here we demonstrate that Tim-1 is a positive regulator of mast cell activation and cytokine production. Similar to our findings on the effects of Tim-3 on mast cells18, Tim-1 expression alone could promote IgE/Ag-mediated NF-κB and NF-AT/AP1 transcriptional activation, without additional cross-linking antibodies or exogenous ligands. Tim-4 is a ligand for Tim-1, but the lack of Tim-4 expression on mast cells makes it an unlikely explanation for this particular role of Tim1 in mast cells. Tim-1 can also bind PS on apoptotic cells24 or PS transiently exposed on degranulating mast cells, either of which could potentially contribute to enhance Tim-1 signaling, although whether PS binding to Tim-1 can lead to mast cell activation is still unknown. Tim-1 has also been reported to bind LMIR5/CD300b, a DAP12-coupled activating receptor expressed on myeloid cells25. Thus, stimulation with TIM1-Fc was able to induce LMIR5-mediated ERK activation in mast cells, suggesting that LMIR5 is another potential endogenous ligand of Tim-1, driving the enhancement of transcriptional response. Finally, Tim-1 may homodimerize through its heavily glycosylated mucin domain, leading to its phosphorylation and downstream function. Regardless of its mode of activation, we showed that Tim-1 co-stimulation is dependent on the tyrosine phosphorylation motif of Tim-1 cytoplasmic tail as mutation of tyrosine 276 rendered Tim-1 unable to mediate its co-stimulatory function. The Src kinase Fyn has been shown to phosphorylate Tim-1 in a B cell line26. We showed that the Tim-1 cytoplasmic tail is phosphorylated upon TCR stimulation in an Lck-dependent manner and can recruit p85 binding8. Therefore, Src family kinases like Lyn, Fyn or Hck are potential facilitators of Tim-1 phosphorylation upon IgE/Ag activation.\n\nAnother aim of this study was to determine how Tim-1 cross-linking by antibodies or ligands could modulate mast cell activation. Unlike in T cells or in vivo experiments, mast cells did not respond to Tim-1 antibody treatment, as none of the antibodies tested elicited a change in mast cell degranulation or cytokine production. On the other hand, Tim-4 treatment consistently enhanced IgE/Ag-mediated cytokine production but not degranulation, which may be a time- and/or signal intensity-dependent effect. Tim-4 was reported to have bimodal effects, either enhancing or inhibiting T cell proliferation, depending on anti-CD3/CD28 concentrations10. This bimodal regulation was later reported to be inhibitory for naïve T cells, which do not express Tim-1, and co-stimulatory for activated T cells, suggesting that Tim-4 either binds to an unknown ligand expressed only on naïve T cells or that Tim-4 has a higher affinity for Tim-1 expressed on activated T cells27. In addition, Tim-4 can bind to naïve T cells that do not express Tim-1 and inhibit Th17 differentiation28. This effect was shown to be independent of Tim-1 activity, since addition of Tim-1 blocking antibody, presumably to block the Tim1:Tim4 interaction, could not rescue Tim4-mediated inhibition28. In contrast to such ligand-dependent effects observed on T cells, Tim-4 co-stimulates IgE/Ag-mediated mast cell activation by cross-linking Tim-1. Using low and high valency antigens to engage the positive and negative signaling pathways of FcεRI, respectively, we showed that Tim-4 could enhance mast cell cytokine production in both settings. Specifically, Tim-4 co-stimulated cytokine release in a dose-dependent manner, under both high and low valency antigen stimulation. These results imply that Tim-4 contributes to FcεRI signaling intensity and/or duration, and may potentially override negative feedback signals linked primarily to Lyn-mediated phosphatase activation. Together with findings that Tim-4 alone does not induce cytokine production in mast cells, our results demonstrate that Tim-1 signaling interfaces with common effector molecules downstream of FcεRI signaling, rather than acting through a parallel pathway, to enhance mast cell functions.\n\nTo determine whether Tim-1 plays a positive or negative regulatory role in mast cells, we first attempted siRNA-mediated knockdown of Tim-1 protein in BMMCs but were unsuccessful in obtaining efficient Tim-1 reduction. Furthermore, two separate strains of Tim-1-deficient mice showed relatively unaltered IgE production and AHR development in an OVA-induced mouse model of asthma, even though one study did observe higher type 2 and Th17 cytokine production in Tim-1 knockout (KO) mice26,29. On the other hand, the importance of the Tim-1 mucin domain has been demonstrated in T cell activation, differentiation, trafficking, and effector function in autoimmunity and airway inflammation28,30. It is also essential to regulatory B cell maintenance, signaling, transplant tolerance and induction of systemic autoimmunity14,15,21. Thus, we examined whether the Tim-1 mucin domain regulates mast cell activity, particularly in the context of Tim-4 treatment. Contrary to the effects seen in B and T cells, the Tim-1 mucin domain is dispensable for mast cell activity, as mast cell degranulation and cytokine release remain intact in the absence of the mucin domain. It is worth noting that Tim-1Δmucin BMMCs were actually able to secrete more cytokine than WT BMMCs in some instances. However, after testing six different batches of WT and Tim-1Δmucin BMMCs, it appeared that any differences observed were due to the maturation status of BMMCs and their FcεRI surface expression, rather than any direct effect of deletion of the Tim-1 mucin domain. While our study focused on bone marrow-derived mast cells, absence of the Tim-1 mucin domain may nonetheless affect trafficking and/or differentiation of other mast cell types in their respective tissue microenvironments in vivo. Tim1-Tim4 interaction is thought to occur mostly through the IgV domains of the respective proteins, although there is evidence that Tim-4 may also bind to the Tim-1 mucin domain10. We showed that Tim-4 mediated co-stimulation of mast cell function occurred independent of the mucin domain. It remains to be determined, in the absence of the Tim-1 mucin domain, whether Tim-4 has other unknown ligands on mast cells that can mediate this enhancement.\n\nWe also examined the signaling pathways leading to enhanced transcriptional activation and cytokine production by Tim-1 and Ag co-stimulation in mast cells. Phosphorylation of Syk was not altered by Tim-4 treatment at the time points we examined. This is consistent with our finding that the Tim1-Tim4 interaction did not promote Ag-stimulated FcεRI signaling intensity, using mast cells from a Nur77GFP mouse model18. Syk is phosphorylated on multiple tyrosines by either auto-phosphorylation or trans-phosphorylation by Lyn, resulting in positive or negative regulation, respectively, of Ag-mediated FcεRI signaling31–33. Since we only examined tyrosines 519 and 520 in the kinase loop, which are sites of Syk auto-phosphorylation, it is possible that other tyrosine phosphorylation sites may be affected by Tim-1. One particular site of interest is tyrosine 346 in the linker region, which is important for PLC-γ1 binding and phosphorylation34. Aside from positive signaling pathways, the relevant negative signaling pathways should be investigated for potential down-regulation by Tim-1 engagement. Thus, a particular FcεRI proximal phosphatase of interest is the Src homology-2-containing signaling protein (SHIP), which has been implicated in regulation of IgE/Ag-induced IL-6 production through inhibition of NF-κB activity, both of which are enhanced by Tim-1 cross-linking in our study35.\n\nIn a previous study, Tim-4/CD3/CD28-coated beads enhanced phosphorylation of Erk and Akt in CD4 T cells9. These results were not observed in previously study using both BMMCs and peritoneal mast cells (PMCs)16. Similarly, we did not observe similar effect with soluble Tim-4 addition in Ag-mediated FcεRI aggregation at the peak of antigen stimulation or when signals returned to basal. It is possible that Tim-4 signals are cell-type specific or that mast cells require more robust aggregation of Tim-1 at the time of antigen stimulation to induce substantial effects. While phosphorylation of Akt was not affected, ribosomal protein S6 was significantly enhanced upon Tim-4 treatment. Specifically, Tim-1 cross-linking enhanced and sustained IgE/Ag-induced S6 phosphorylation, which correlates with enhanced cytokine production. Ribosomal protein S6 is a downstream effector of PI3K/mTOR signaling, and as such is essential for mast cell survival, proliferation, metabolism and protein synthesis. MALT1 and BCL10, members of the Carma1-MALT1-Bcl10 (CBM) complex, are essential regulators of FcεRI-induced mast cell activation by selectively up-regulating NF-κB-dependent cytokine production without affecting degranulation and leukotriene synthesis36. We previously identified a Carma1-MALT1-dependent activation of mTOR signaling after TCR engagement, leading to phosphorylation of S6 and another mTOR substrate 4E-BP137. Thus, Tim-1 may preferentially engage the MALT1-Bcl10 pathway to modulate mTOR signaling and NF-κB responses without affecting FcεRI-proximal signaling. Consequently, further studies are needed to address whether Tim1-Tim4 interaction promotes mast cell metabolic responses and protein synthesis as well as whether the specific Tim-1 targets in this pathway leading to effector functions.\n\nTaken together, our findings provide further evidence that Tim-1 signaling can promote cytokine production in IgE/Ag-activated mast cells. This is in line with a co-stimulatory role for Tim-1 in T, B and NKT cells, both in vitro and in vivo. Contrary to their previously described effects on these cell types, Tim-1 antibodies did not regulate mast cell degranulation and cytokine production in our hands. Nevertheless, Tim-1 ligation by Tim-4 consistently enhanced mast cell cytokine production, and this effect was not affected by loss of the Tim-1 mucin domain. We also showed, for the first time, that unlike in regulatory B cells, the Tim-1 mucin domain is dispensable for mast cell effector function. Mast cells play diverse roles in hypersensitivity, allergic and, as has been increasingly appreciated, non-allergic diseases17. Tim-1 blockade or cross-linking by known ligands and antibodies has been shown to ameliorate or exacerbate allergic lung inflammation in vivo. Therefore, manipulation of Tim-1 activity on mast cells, particularly modulation of the Tim1-Tim4 interaction, could be a novel therapeutic target to control allergic and autoimmune disease.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for Figure 1–Figure 5 in ‘Mast cell activation is enhanced by Tim1:Tim4 interaction but not by Tim-1 antibodies’, 10.5256/f1000research.8132.d11487040",
"appendix": "Author contributions\n\n\n\nLPK conceived the study. BLP and LPK designed the experiments. All experiments were carried out by BLP. Data analysis and interpretation were performed by BLP and LPK. BLP wrote the manuscript, with editorial assistance from LPK.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by PHS grant R56AI067544 to LPK.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank V. Kuchroo for providing Tim-1 antibodies and Tim4-Fc recombinant protein. We also thank V. Kuchroo and D. Rothstein for the Tim-1Δmucin mice.\n\n\nReferences\n\nFeigelstock D, Thompson P, Mattoo P, et al.: The human homolog of HAVcr-1 codes for a hepatitis A virus cellular receptor. J Virol. 1998; 72(8): 6621–8. PubMed Abstract | Free Full Text\n\nMcIntire JJ, Umetsu SE, Akbari O, et al.: Identification of Tapr (an airway hyperreactivity regulatory locus) and the linked Tim gene family. Nat Immunol. 2001; 2(12): 1109–16. PubMed Abstract | Publisher Full Text\n\nLee J, Phong B, Egloff AM, et al.: TIM polymorphisms--genetics and function. Genes Immun. 2011; 12(8): 595–604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcIntire JJ, Umetsu SE, Macaubas C, et al.: Immunology: hepatitis A virus link to atopic disease. Nature. 2003; 425(6958): 576. PubMed Abstract | Publisher Full Text\n\nUmetsu SE, Lee WL, McIntire JJ, et al.: TIM-1 induces T cell activation and inhibits the development of peripheral tolerance. Nat Immunol. 2005; 6(5): 447–54. PubMed Abstract | Publisher Full Text\n\nde Souza AJ, Oriss TB, O'Malley K J, et al.: T cell Ig and mucin 1 (TIM-1) is expressed on in vivo-activated T cells and provides a costimulatory signal for T cell activation. Proc Natl Acad Sci U S A. 2005; 102(47): 17113–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSizing ID, Bailly V, McCoon P, et al.: Epitope-dependent effect of anti-murine TIM-1 monoclonal antibodies on T cell activity and lung immune responses. J Immunol. 2007; 178(4): 2249–61. PubMed Abstract | Publisher Full Text\n\nde Souza AJ, Oak JS, Jordanhazy R, et al.: T cell Ig and mucin domain-1-mediated T cell activation requires recruitment and activation of phosphoinositide 3-kinase. J Immunol. 2008; 180(10): 6518–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez-Manzanet R, Meyers JH, Balasubramanian S, et al.: TIM-4 expressed on APCs induces T cell expansion and survival. J Immunol. 2008; 180(7): 4706–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyers JH, Chakravarti S, Schlesinger D, et al.: TIM-4 is the ligand for TIM-1, and the TIM-1-TIM-4 interaction regulates T cell proliferation. Nat Immunol. 2005; 6(5): 455–64. PubMed Abstract | Publisher Full Text\n\nBinné LL, Scott ML, Rennert PD: Human TIM-1 associates with the TCR complex and up-regulates T cell activation signals. J Immunol. 2007; 178(7): 4342–50. PubMed Abstract | Publisher Full Text\n\nHan WK, Bailly V, Abichandani R, et al.: Kidney Injury Molecule-1 (KIM-1): a novel biomarker for human renal proximal tubule injury. Kidney Int. 2002; 62(1): 237–44. PubMed Abstract | Publisher Full Text\n\nKim HY, Chang YJ, Chuang YT, et al.: T-cell immunoglobulin and mucin domain 1 deficiency eliminates airway hyperreactivity triggered by the recognition of airway cell death. J Allergy Clin Immunol. 2013; 132(2): 414–25.e6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYeung MY, Ding Q, Brooks CR, et al.: TIM-1 signaling is required for maintenance and induction of regulatory B cells. Am J Transplant. 2015; 15(4): 942–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing Q, Yeung M, Camirand G, et al.: Regulatory B cells are identified by expression of TIM-1 and can be induced through TIM-1 ligation to promote tolerance in mice. J Clin Invest. 2011; 121(9): 3645–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakae S, Iikura M, Suto H, et al.: TIM-1 and TIM-3 enhancement of Th2 cytokine production by mast cells. Blood. 2007; 110(7): 2565–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nda Silva EZ, Jamur MC, Oliver C: Mast cell function: a new vision of an old cell. J Histochem Cytochem. 2014; 62(10): 698–738. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhong BL, Avery L, Sumpter TL, et al.: Tim-3 enhances FcεRI-proximal signaling to modulate mast cell activation. J Exp Med. 2015; 212(13): 2289–304. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiao W, Nishimoto H, Hong H, et al.: Positive and negative regulation of mast cell activation by Lyn via the FcepsilonRI. J Immunol. 2005; 175(10): 6885–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiao S, Najafian N, Reddy J, et al.: Differential engagement of Tim-1 during activation can positively or negatively costimulate T cell expansion and effector function. J Exp Med. 2007; 204(7): 1691–702. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiao S, Brooks CR, Zhu C, et al.: Defect in regulatory B-cell function and development of systemic autoimmunity in T-cell Ig mucin 1 (Tim-1) mucin domain-mutant mice. Proc Natl Acad Sci U S A. 2012; 109(30): 12105–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKambayashi T, Koretzky GA: Proximal signaling events in Fc epsilon RI-mediated mast cell activation. J Allergy Clin Immunol. 2007; 119(3): 544–52; quiz 53–4. PubMed Abstract | Publisher Full Text\n\nKim MS, Kuehn HS, Metcalfe DD, et al.: Activation and function of the mTORC1 pathway in mast cells. J Immunol. 2008; 180(7): 4586–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKobayashi N, Karisola P, Peña-Cruz V, et al.: TIM-1 and TIM-4 glycoproteins bind phosphatidylserine and mediate uptake of apoptotic cells. Immunity. 2007; 27(6): 927–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYamanishi Y, Kitaura J, Izawa K, et al.: TIM1 is an endogenous ligand for LMIR5/CD300b: LMIR5 deficiency ameliorates mouse kidney ischemia/reperfusion injury. J Exp Med. 2010; 207(7): 1501–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCurtiss ML, Hostager BS, Stepniak E, et al.: Fyn binds to and phosphorylates T cell immunoglobulin and mucin domain-1 (Tim-1). Mol Immunol. 2011; 48(12–13): 1424–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMizui M, Shikina T, Arase H, et al.: Bimodal regulation of T cell-mediated immune responses by TIM-4. Int Immunol. 2008; 20(5): 695–708. PubMed Abstract | Publisher Full Text\n\nCao W, Ryan M, Buckley D, et al.: Tim-4 inhibition of T-cell activation and T helper type 17 differentiation requires both the immunoglobulin V and mucin domains and occurs via the mitogen-activated protein kinase pathway. Immunology. 2011; 133(2): 179–89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarlow JL, Wong SH, Ballantyne SJ, et al.: Tim1 and Tim3 are not essential for experimental allergic asthma. Clin Exp Allergy. 2011; 41(7): 1012–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAngiari S, Donnarumma T, Rossi B, et al.: TIM-1 glycoprotein binds the adhesion receptor P-selectin and mediates T cell trafficking during inflammation and autoimmunity. Immunity. 2014; 40(4): 542–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang J, Kimura T, Siraganian RP: Mutations in the activation loop tyrosines of protein tyrosine kinase Syk abrogate intracellular signaling but not kinase activity. J Immunol. 1998; 161(8): 4366–74. PubMed Abstract\n\nSanderson MP, Gelling SJ, Rippmann JF, et al.: Comparison of the anti-allergic activity of Syk inhibitors with optimized Syk siRNAs in FcepsilonRI-activated RBL-2H3 basophilic cells. Cell Immunol. 2010; 262(1): 28–34. PubMed Abstract | Publisher Full Text\n\nSada K, Zhang J, Siraganian RP: Point mutation of a tyrosine in the linker region of Syk results in a gain of function. J Immunol. 2000; 164(1): 338–44. PubMed Abstract | Publisher Full Text\n\nLaw CL, Chandran KA, Sidorenko SP, et al.: Phospholipase C-gamma1 interacts with conserved phosphotyrosyl residues in the linker region of Syk and is a substrate for Syk. Mol Cell Biol. 1996; 16(4): 1305–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalesnikoff J, Baur N, Leitges M, et al.: SHIP negatively regulates IgE + antigen-induced IL-6 production in mast cells by inhibiting NF-kappa B activity. J Immunol. 2002; 168(9): 4737–46. PubMed Abstract | Publisher Full Text\n\nKlemm S, Gutermuth J, Hültner L, et al.: The Bcl10-Malt1 complex segregates Fc epsilon RI-mediated nuclear factor kappa B activation and cytokine production from mast cell degranulation. J Exp Med. 2006; 203(2): 337–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamilton KS, Phong B, Corey C, et al.: T cell receptor-dependent activation of mTOR signaling in T cells is mediated by Carma1 and MALT1, but not Bcl10. Sci Signal. 2014; 7(329): ra55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDemo SD, Masuda E, Rossi AB, et al.: Quantitative measurement of mast cell degranulation using a novel flow cytometric annexin-V binding assay. Cytometry. 1999; 36(4): 340–8. PubMed Abstract | Publisher Full Text\n\nKane LP, Shapiro VS, Stokoe D, et al.: Induction of NF-kappaB by the Akt/PKB kinase. Curr Biol. 1999; 9(11): 601–4. PubMed Abstract | Publisher Full Text\n\nPhong B, Kane L: Dataset 1 in: Mast cell activation is enhanced by Tim1:Tim4 interaction but not by Tim-1 antibodies. F1000Research. 2016. Data Source"
}
|
[
{
"id": "12697",
"date": "15 Mar 2016",
"name": "Avery August",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere are a few issues that the authors should address as they refine the presentation of this work as detailed below:The authors should refer to the data in Figure 1A, the characterization of the TIM-4-Fc reagent and expression of TIM-1 on the surface of these mast cells. Since IgE alone has been shown to affect mast cell survival, does TIM-4/TIM-1 affect the mast cell response to IgE alone? In figure 1E, the graph looks as if there is production of IL-6 in the absence of IgE (third bar from the left). Is this correct? The location of the flag tag on TIM-1 should be described by the authors. Do the antibodies against TIM-1 affect the ability of TIM-4 to enhance cytokine production? The reference to a small although statistically insignificant increase in IL6 when 4G8 was used is not appropriate to say, since the change is not significant, and not any larger than other changes that are not remarked upon by the authors. This statement should be modified. In experiments where TIM-1 is overexpressed in the MC/9 cell line, what is the ligand for the transfected TIM-1? Is calcium signaling affected by TIM-4/Tim-1 interaction? Is activation of Btk affected given the potential role of PI3K downstream of TIM-1? The statement in the 6th line of the discussion starting with “Tim-1 expression alone could promote IgE/Ag-mediated NF-kB and NF-AT/AP1 transcriptional activation…” should be qualified that this is in MC/9 cells, which may not behave like primary mast cells.",
"responses": [
{
"c_id": "2046",
"date": "08 Jul 2016",
"name": "Larry Kane",
"role": "Author Response F1000Research Advisory Board Member",
"response": "The authors should refer to the data in Figure 1A, the characterization of the TIM-4-Fc reagent and expression of TIM-1 on the surface of these mast cells. This is now added to the Results, with an extra description of the Tim4-Fc in the methods Since IgE alone has been shown to affect mast cell survival, does TIM-4/TIM-1 affect the mast cell response to IgE alone? We did not test the IgE+Tim4-Fc condition ourselves but a previous paper from Galli and colleagues (ref. 16 in the manuscript) showed that IgE + rmTim-4 did not trigger degranulation, cytokine production, or prevent mast cell apoptosis from IL-3 withdrawal. In figure 1E, the graph looks as if there is production of IL-6 in the absence of IgE (third bar from the left). Is this correct? That condition actually contains both IgE and the indicated antigen; the bar to the left of it represents cells receiving only IgE and nothing else. The location of the flag tag on TIM-1 should be described by the authors. The tag is at the N-terminus. This is now stated explicitly in the Results section. Do the antibodies against TIM-1 affect the ability of TIM-4 to enhance cytokine production? We did not perform any blocking experiments with the Tim-1 antibodies in the presence of Tim-4 fusion protein, but this would be an interesting future experiment. The reference to a small although statistically insignificant increase in IL6 when 4G8 was used is not appropriate to say, since the change is not significant, and not any larger than other changes that are not remarked upon by the authors. This statement should be modified. This is a good point. We have now modified the sentence in question which refers to Fig. 2F in the results section to make it more accurately and clearly reflect the data. In experiments where TIM-1 is overexpressed in the MC/9 cell line, what is the ligand for the transfected TIM-1? In our previous publications, we observed what appears to be constitutive activity of the transfected Tim-1 after transient transfection. We have generally interpreted this to indicate that ectopically expressed Tim-1 is oligomerizing in the absence of ligand, since it is expressed at higher than normal levels on at least a subset of the transfected cells. Is calcium signaling affected by TIM-4/Tim-1 interaction? Is activation of Btk affected given the potential role of PI3K downstream of TIM-1? Given the data we have presented in the manuscript, it is indeed logical that a Tec family kinase like Btk may be a proximal target for Tim-1. We have started to address this by probing the effects of Tim-1 on PLC-g1 phosphorylation, but have not been able to observe a robust effect. However, given the limited personnel available to work currently on this project, we thought that it was important to share the results that we have obtained, which may be of interest to others studying the function of Tim-1. The statement in the 6th line of the discussion starting with “Tim-1 expression alone could promote IgE/Ag-mediated NF-kB and NF-AT/AP1 transcriptional activation…” should be qualified that this is in MC/9 cells, which may not behave like primary mast cells. This is a good point. We have now clarified in the Discussion that this was in MC/9 cells."
}
]
},
{
"id": "14555",
"date": "23 Jun 2016",
"name": "Bridget S Wilson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe TIM-1 protein has been implicated in regulation of T cells, NKT cells and regulatory B cells.\n\nHere, the focus is on mast cells, where Tim-1 has been shown to be constitutively expressed and is believed to have a co-stimulatory function for NF-kB and NFAT/AP1 transcription and cytokine production.\n\nA few points would be helpful to address in the final version of the manuscript.\nPrior work has shown that Tim-1 recognizes phosphatidylserine on apoptotic cells and that the mucin domain is important for this binding. Since the data presented here show that annexin-binding (which detects PS) is upregulated as a function of mast cell degranulation, the Tim-1 mucin domain might mediate homotypic interactions between recently degranulated cells. Is there evidence that this occurs (microscopy images or flow-based measures?) If this is a mechanism to cluster Tim1 for activation, it will be in a “synapse”-like setting and would support a post-degranulation, delayed co-stimulatory signal. As mentioned below, Tim-4 mediated activation in vivo is probably in a synapse setting too, so there seems like a missed opportunity here.\n\nThis study relies a lot on exogenous agents (Tim4-Fc, antibodies) to make conclusions about the role of Tim-1 signaling. This is fine as an experimental approach, but can confound the conclusions reached about the role of the mucin domain. In other words, if one bypasses the mucin domain by activating with Tim4-Fc, it does not necessarily mean that ligations mediated via the mucin domain does not mediate signaling in another setting. I find, in general, that the emphasis on this result is overstated/over-discussed as a physiological relevant finding. It provides more insight on the structural features of Tim-1, which can apparently be activated by crosslinking (ie antibodies) or by binding to another Tim (Tim4) presented on another cell… and possibly when PS, presented on another mast cell or an apoptotic cell in the inflammatory microenvironment, clusters Tim-1 via the mucin domain.\nIn keeping with this, I think the discussion could be shortened and simplified. For example, this text could be omitted or reworded:\n“Thus, we examined whether the Tim-1 mucin domain regulates mast cell activity, particularly in the context of Tim-4 treatment. Contrary to the effects seen in B and T cells, the Tim-1 mucin domain is dispensable for mast cell activity, as mast cell degranulation and cytokine release remain intact in the absence of the mucin domain. It is worth noting that Tim-1 [] mucin BMMCs were actually able to secrete more cytokine than WT BMMCs in some instances. However, after testing six different batches of WT and Tim-1mucin BMMCs, it appeared that any differences observed were due to the maturation status of BMMCs and their Fc[epsilon]RI surface expression, rather than any direct effect of deletion of the Tim-1 mucin domain.”\n\nThis group has opted to use SPE-7 (anti-DNP) IgE to prime the BMMC. Since this is a cytokinergic IgE, it could be good to confirm that the essential findings of co-stimulation also occur when using a non-cytokinergic (anti-DNP) IgE …perhaps the Liu IgE.",
"responses": [
{
"c_id": "2047",
"date": "08 Jul 2016",
"name": "Larry Kane",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Prior work has shown that Tim-1 recognizes phosphatidylserine on apoptotic cells and that the mucin domain is important for this binding. Since the data presented here show that annexin-binding (which detects PS) is upregulated as a function of mast cell degranulation, the Tim-1 mucin domain might mediate homotypic interactions between recently degranulated cells. Is there evidence that this occurs (microscopy images or flow-based measures?) If this is a mechanism to cluster Tim1 for activation, it will be in a “synapse”-like setting and would support a post-degranulation, delayed co-stimulatory signal. As mentioned below, Tim-4 mediated activation is probably in a synapse setting too, so there seems like a missed opportunity here. The reviewer raises an interesting point regarding the regulation of Tim-1 by PS on degranulating mast cells (although to our knowledge, the Tim1:PS interaction mainly occurs through the Tim-3 IgV domain). We have not yet been able to image Tim-1 during mast cell activation, but feel that this would be an interesting topic for future exploration. This study relies a lot on exogenous agents (Tim4-Fc, antibodies) to make conclusions about the role of Tim-1 signaling. This is fine as an experimental approach, but can confound the conclusions reached about the role of the mucin domain. In other words, if one bypasses the mucin domain by activating with Tim4-Fc, it does not necessarily mean that ligations mediated via the mucin domain does not mediate signaling in another setting. I find, in general, that the emphasis on this result is overstated/over-discussed as a physiological relevant finding. It provides more insight on the structural features of Tim-1, which can apparently be activated by crosslinking (ie antibodies) or by binding to another Tim (Tim4) presented on another cell… and possibly when PS, presented on another mast cell or an apoptotic cell in the inflammatory microenvironment, clusters Tim-1 via the mucin domain. This is a fair point.We have modified the discussion of these results in line with this comment and the one below. In keeping with this, I think the discussion could be shortened and simplified. For example, this text could be omitted or reworded: “Thus, we examined whether the Tim-1 mucin domain regulates mast cell activity, particularly in the context of Tim-4 treatment. Contrary to the effects seen in B and T cells, the Tim-1 mucin domain is dispensable for mast cell activity, as mast cell degranulation and cytokine release remain intact in the absence of the mucin domain. It is worth noting that Tim-1 d.mucin BMMCs were actually able to secrete more cytokine than WT BMMCs in some instances. However, after testing six different batches of WT and Tim-1 d.mucin BMMCs, it appeared that any differences observed were due to the maturation status of BMMCs and their Fc[epsilon]RI surface expression, rather than any direct effect of deletion of the Tim-1 mucin domain.” This is a good suggestion – we have now edited the Discussion, to make it more readable, and as a result it has been shortened by about one page. This group has opted to use SPE-7 (anti-DNP) IgE to prime the BMMC. Since this is a cytokinergic IgE, it could be good to confirm that the essential findings of co-stimulation also occur when using a non-cytokinergic (anti-DNP) IgE …perhaps the Liu IgE. We have only used the commercially available, and widely employed,SPE-7 IgE in our studies. We agree that branching out to other IgE’s would be an interesting avenue for future follow-up exploration."
}
]
}
] | 1
|
https://f1000research.com/articles/5-251
|
https://f1000research.com/articles/5-563/v1
|
05 Apr 16
|
{
"type": "Method Article",
"title": "Detecting variable responses within fMRI time-series of volumes-of-interest using repeated measures ANOVA",
"authors": [
"Paul M. Macey",
"Philip J. Schluter",
"Katherine E. Macey",
"Ronald M. Harper",
"Philip J. Schluter",
"Katherine E. Macey",
"Ronald M. Harper"
],
"abstract": "We present an approach to analyzing fMRI timetrends from volumes-of-interest (VOI) within and between subject groups using repeated measures analysis of variance (RMANOVA), which allows temporal patterns to be examined without an a priori model of expected timing or pattern of response. The method serves as a complement to whole-brain voxel-based analyses, and is useful for detecting complex responses within pre-determined brain regions, or as a post-hoc analysis of regions of interest identified by whole-brain assessments. We illustrate an implementation of the technique in the statistical software package SAS. VOI timetrends are extracted from conventionally preprocessed fMRI images. A timetrend of average signal intensity across the VOI during the scanning period is calculated for each subject. The values are scaled relative to baseline periods, imported into SAS, and the procedure PROC MIXED implements the RMANOVA. The ensuing results allow determination of significant overall effects, and time-point specific within- and between-group responses relative to baseline. We illustrate the technique using fMRI data from two groups of subjects who underwent a respiratory challenge. RMANOVA allows insight into the timing of responses and response differences between groups, and so is suited to fMRI paradigms eliciting complex response patterns.",
"keywords": [
"mixed effect models",
"regression analysis",
"statistical models",
"physiological responses",
"functional magnetic resonance imaging"
],
"content": "Introduction\n\nWe describe a procedure for analyzing fMRI time-courses across multiple subjects in pre-determined regions without an a priori pattern or timing of expected response. Functional MRI analyses usually require defining a model and identifying voxels whose time courses follow that model; other model-free event-related approaches test for responses that are coincident in time or duration. However, certain tasks elicit neural responses that vary in pattern and timing. For example, responses to physiological challenges typically occur across neural structures in a sequence rather than simultaneously, and the patterns differ according to structure1–3. While existing software like SPM does allow for model-free assessment of responses, such as identifying the hemodynamic response based on a finite impulse response function in SPM124, those approaches can be complex to implement and interpret. The present method arose out of our research group’s need to perform basic time-trend analyses of fMRI signals in specific regions.\n\nThe approach we propose is analyzing fMRI time series using repeated measures analysis of variance (RMANOVA), allowing for detection of within-group responses relative to a baseline or resting state, and of between-group differences in response. As with any analysis of variance (ANOVA), RMANOVA tests the equality of means that are assumed to be normally distributed. However, RMANOVA also accounts for the correlation between repeated measures within subjects, whereas ANOVA does not. In addition to assessment of significant reactions relative to a baseline period, or of significant group differences in response, the method allows for identifying specific time-points when these differences occur. Therefore, transient, late-developing, and complex sequences of increasing and decreasing signal changes can be distinguished with RMANOVA. The technique is suited to analysis of longer task paradigms evoking a complex pattern of responses. We present an implementation of RMANOVA in the programming language SAS (SAS Institute Inc., Cary, NC).\n\n\nDescription of procedure\n\nFunctional MRI data collected in one or more groups of subjects are preprocessed, typically with slice timing and motion correction, and the images may be spatially normalized. An additional, optional step is detrending to remove global effects. In traditional cluster analyses, global effects may be accounted for within the analysis steps using, for example, SPM5. Intensity normalization using proportional scaling has been the most common approach to removing such effects, but for the example presented here, removal of global signal changes is performed on the preprocessed images prior to analysis with RMANOVA using a custom technique6.\n\nVolumes-of-interest (VOI) are defined, usually by outlining a structure of interest on an image volume and saving that outline as a binary image, or mask. The VOI may be defined individually on a subject-by-subject basis, or globally for all subjects on a template (the latter case assumes subjects’ images are accurately spatially normalized to the template). The mask is used to extract the intensity of only those voxels within the VOI, at each time point in the fMRI series. For each time point, the average of those voxel values is recorded, resulting in a time-series corresponding to the blood oxygen level dependent (BOLD) signal time course of that VOI in the fMRI series. The procedure is repeated for all subjects, resulting in one time trend for each subject per VOI. The time trends may optionally be deconvolved with a hemodynamic response function (HRF), allowing the timing of significant responses to be inferred as neural, rather than BOLD. The resulting VOI time trends are subjected to RMANOVA.\n\nRepeated measures ANOVA can be implemented using a mixed effect model, a generalization of the standard linear model which allows for correlation and non-constant variability within the data7. The methodology described here assumes use of the SAS function PROC MIXED, although other software packages also perform RMANOVA. However, SAS is very flexible, and includes a wide range of diagnostics tests. (See white paper “Comparing the SAS® GLM and MIXED Procedures for Repeated Measurements Analysis,” Russ Wolfinger and Ming Chang, SUGI Proceedings, 1995; available at https://support.sas.com/rnd/app/stat/papers/abstracts/mixedglm.html). The statistical procedures calculate the variance-covariance matrix based on the dependencies defined by repeated measurements within subjects, and by group classification of subjects in the case of more than one group (REPEATED and GROUP options within PROC MIXED; by convention, SAS syntax is in upper case)7.\n\nA classification variable (t) representing baseline and time-points in the task or challenge period is created, with values of t during the baseline or reference period constant, and values increasing for each time point in the series. For example, we have collected fMRI data during physiological challenges with one minute of baseline followed by 90 sec of challenge, at a repetition time of 6 seconds (25 volumes total); thus, t is 0 for the first 10 scans, and 1 for the 11th volume, 2 for the 12th, and so on until 15 for the 25th. Note that for data collected at higher sampling rates or across longer time periods, the time-classification variable may include multiple data points, as opposed to the individual data points of the fMRI time-series; for example, we have classified instantaneous (beat-by-beat) heart rates recorded throughout a two minute period into twelve 10 sec epochs8. If more than one group of subjects is being analyzed, a second classification variable is defined at each time point with a value representing the group to which that series belongs. An example of such a format in an Excel file is shown in Appendix A; these data may be imported into SAS.\n\nBefore assessing the fMRI time-trend results of the RMANOVA analysis (see sections below for full details), certain residual and influence diagnostic tests should be performed to ensure that the model does not seriously violate underlying assumptions. A relevant subset of such tests is described in sections below. Many practical applications of RMANOVA violate the model assumptions to some degree, but the approach is considered robust to moderate departures of these assumptions. In the present application to fMRI data, a key test is to ensure that the residual distributions are not seriously skewed. We suggest the flowing checks. First determine the s residuals, which are scaled so that more than 95% should fall within the critical limits of the appropriate Student’s t distribution, which for most purposes can be assumed to be within ± 1.96 (and centered on 0). The shape of the histogram for each group of the studentized residuals should be bell-shaped. There should also be no effect of time, so each group plot of the studentized residuals by time should not show a trend. We recommend using locally-weighted-scatter-plot (lowess) curves to identify such trends. In the case where the diagnostic tests fail, there may still be some value in continuing with the model, but we suggest presenting the results of the residual diagnostics along with the findings, to allow readers to assess for themselves the validity of the model. Further options are explained in sections below, along with other possible diagnostic tests.\n\nThere are two levels of results produced by RMANOVA: the significance of the overall model at the group, time, and group-by-time levels, and the significance of between-group and within-group effects at each time point (see section below). According to the Tukey-Fisher criterion for multiple comparisons, the model is first assessed at the overall level for significance before investigating time-point specific effects; i.e., if the overall group-by-time effect is not significant, the between-group effects at each time point are not considered, regardless of their reported significance level. Similarly, if the overall time effect is not significant, then within-group effects at each time point are not considered.\n\nSpecific programming details are presented in the next sections to facilitate replication.\n\nData are arranged in rows per subject. Each subject is identified by a unique classification variable (SUBJECT, which may be the subject name), and similarly, the group to which that subject belongs is designated by the GROUP classification variable e.g., 1, 2, …etc.). The fMRI times series values are arranged in adjacent columns, with one column for each time point (image volume), and with the value representing the average of the voxels within the VOI at that time point for that subject (row in Figure 1). Values are in percentage change relative to a baseline period. The column headings for the time points take the format of a root (“val” in Figure 1 and the example code in next section) followed by a number indicating the sequence of values, starting at 0 (e.g., val0, val1, val2, …etc.).\n\nThe SAS code for importing these data into a library “voilib” is as follows:\n\nLIBNAME voilib ‘[sas_library_folder]’;\n\nPROC IMPORT OUT=voilib.data\n\nDATAFILE = '[fn]'\n\nDBMS=EXCEL2000 REPLACE;\n\nRANGE = '[worksheet]'\n\nGETNAMES=YES;\n\nRUN;\n\nwhere [sas_library_folder] is the path to where the SAS library files will be stored, [fn] is the Excel file name, and [worksheet] is the name of the worksheet within the Excel file containing the data.\n\nSAS can also import data from a text file, which is useful for series with more that 250 time points, or for systems without Excel. The SAS help has details on importing such files.\n\nNote that SAS code is not case dependent. In the examples presented here, capitals are used to indicate SAS commands, and bolded names within square brackets are used to indicate user-specified variables. The remaining lowercase names are variables specific to these examples.\n\nA simple example of formatting the data is illustrated with the following SAS code, where a variable “epoch” classifies all time points as baseline epoch or epoch number corresponding to task scan number, with the assumption that one baseline period is followed by the task period:\n\nPROC FORMAT;\n\nVALUE ynft 1='yes' 0='no';\n\nVALUE gft 1='[group1]' 2 = '[group2]';\n\nRUN;\n\nDATA voilib.formated; SET voilib.data;\n\nARRAY y{[num_tp]} val0 – val [num_tp-1];\n\nDO i=1 TO [num_tp];\n\ntime=i; val=y{i};\n\nIF 0 < time <= [bl] THEN epoch=[bl];\n\nIF time > [bl] THEN epoch=time;\n\nIF 0 < time <= [bl] THEN chal=’baseline’;\n\nIF time > [bl] THEN chal=’challenge’;\n\nOUTPUT;\n\nEND;\n\nKEEP subject group time epoch chal;\n\nRUN;\n\nwhere [num_tp] is the number of values in the time-series and [bl] is the last baseline scan. The names for each group may be entered in place of [group1] and [group2]. The above formatting code can be modified according to the fMRI task timing. For time series with large numbers of samples, the epoch variable can be used to combine multiple values within a time period.\n\nThe RMANOVA, along with formatting and output filtering, is implemented as follows:\n\nPROC MIXED NOITPRINT NOCLPRINT DATA=voilib.formated EMPIRICAL;\n\nCLASS group epoch subject;\n\nMODEL val = group epoch group*epoch / OUTP = prediction RESIDUAL;\n\nREPEATED / TYPE=CS SUB=subject(group) group=group;\n\nLSMEANS group*epoch / DIFF;\n\nODS LISTING EXCLUDE LSMEANS;\n\nODS OUTPUT LSMEANS= voilib.means;\n\nODS LISTING EXCLUDE DIFFS;\n\nODS OUTPUT DIFFS= voilib.differences;\n\nODS OUTPUT COVPARMS= voilib.covparams;\n\nODS OUTPUT LRT= voilib.likelihoodratiotest;\n\nODS OUTPUT TESTS3= voilib.type3tests;\n\nRUN;\n\nwhere the italicized statements define the model, and the remaining control the output (including residuals). Note that the RESIDUAL option, which directs studentized residuals to be calculated, is only available from SAS version 9.1. The CLASS statement determines classification variables.\n\nThe EMPIRICAL option in the proc mixed command line uses the Huber-White sandwich estimator of variance, rather than the default, as it is more conservative but considered more robust, particularly if the assumption of normality is violated (as is often the case).\n\nThe output from this step contains numerous descriptors of the model. The “Null Model Likelihood Ratio Test” gives the probability of the null model being true (column heading “Pr > ChiSq”). If the model passes the Tukey-Fisher criterion for multiple comparisons, i.e., the probability of the null model is less the 0.05, the model is investigated at the variable level using the “Type 3 Tests of Fixed Effects” output. Probabilities of the null contribution to the model by individual independent variables are presented in the “Pr > F” column for the group, epoch and group*epoch. “Group” represents a group effect, which is often not significant for fMRI data presented in percent change relative to baseline. The variable epoch represents a time effect, corresponding to within-group significant responses. The group*epoch term represents a group-by-time effect, corresponding to between-group differences across time. Again considering the Tukey-Fisher criterion, if for any of these variables the likelihood of the null model is greater than 0.05, then that variable is determined to be not significant. If less than the threshold, the model is investigated further at each time point.\n\nFurther code is needed to select the relevant output for determining the time points of within and between group significant differences. The code below is one of several possible ways to extract the relevant results.\n\nDATA voilib.means; SET voilib.means;\n\nRENAME estimate=adjmean;\n\nDROP effect df tValue Probt;\n\nRUN;\n\nPROC SORT DATA= voilib.means; BY group epoch; RUN;\n\nDATA voilib.differences; SET voilib.differences;\n\nDROP effect df;\n\nRENAME estimate=meandiff;\n\nLABEL estimate='mean diff' _epoch='vs epoch' _group='vs group' probt='p value';\n\nRUN;\n\nDATA voilib.betweengroup; SET voilib.differences;\n\nIF epoch ne _epoch THEN delete;\n\nDROP _epoch;\n\nRUN;\n\nPROC SORT DATA= voilib.betweengroup; BY epoch group; RUN;\n\nDATA voilib.withingroup; SET voilib.differences;\n\nIF group NE _group THEN delete;\n\nDROP _group;\n\nIF epoch NE 28 THEN delete;\n\nRUN;\n\nPROC SORT DATA= voilib.withingroup; BY group _epoch; RUN;\n\nThe above code will create Withingroup and Betweengroup tables in the SAS library voilib.\n\nIf “epoch” showed a significant effect in the Type 3 tests (above), the time points where this difference appeared may be determined by examining the Withingroup table. As shown in Figure 2, each row in the table compares baseline (labeled “epoch”) vs subsequent epochs (labeled “vs epoch”), with a mean difference, standard error and P value; if the latter indicates significance (< 0.05), that time point is considered to have a significant response relative to baseline for that group.\n\nSimilarly, if “group*epoch” showed a significant effect in the Type 3 tests (above), the time-points where this difference appeared may be determined by examining the Betweengroup table. As shown in Figure 3, each row corresponds to a time point (labeled “epoch”) and compares one group vs the other, with a mean difference, standard error and P value; if the latter indicates significance (< 0.05), that time point is considered to have a significant group difference.\n\nThe RMANOVA method has a number of underlying assumptions which should be considered. The assumptions of the RMANOVA method which differ from ANOVA are that 1) the means are linear over time, 2) multivariate normality, 3) homogeneity of covariance matrices, and 4) independence. The method is reasonably robust to violations of the second and third assumptions. Violations of independence produce a non-normal distribution of the residuals, which invalidates the F-ratio. The most common violations of independence occur when either random selection or random assignment is not used or the compound symmetry covariance assumption is inappropriate. Violation of the homogeneity of covariance matrices generally results in the overall test having a higher Type I error rate than nominally set.\n\nIn the current application, a subset of key diagnostic and residual checks is suggested with the implementation of this technique.\n\nPlots of predicted and observed data, to visually determine any major bias.\n\nResidual checks of normality, including 95% of Studentized residuals falling within ±1.96, showing approximately normal group distributions, and scatter plots over time with lowess curves superimposed to ensure lack of group trends with time. Formal tests of normality may also be performed, although these will typically fail.\n\nAssuming the “RESIDUAL” option in PROC MIXED has been used, which implements the studentized residuals, the following SAS code implements the tests:\n\nPROC SORT DATA=voilib.formatted; BY group epoch; RUN;\n\nPROC SORT DATA=voilib.means; BY group epoch; RUN;\n\nPROC SORT DATA=voilib.prediction; BY group epoch; RUN;\n\nDATA voilib.residuals;\n\nMERGE voilib.formatted voilib.means voilib.prediction;\n\nBY group epoch;\n\nresid = val - adjmean;\n\nLABEL resid='residual errors';\n\nabsresid = abs(resid);\n\nRUN;\n\nPROC UNIVARIATE NORMAL PLOT DATA=voilib.residuals;\n\nVAR resid; BY group;\n\nTITLE 'Residual Analysis';\n\nRUN;\n\nSYMBOL1 I=NONE V=STAR; * C=BLACK; *;\n\nPROC GPLOT DATA=voilib.residuals;\n\nPLOT resid*adjmean / VREF=0; BY group;\n\nTITLE 'Residuals';\n\nRUN;\n\nDATA voilib.resid_count; SET voilib.residuals;\n\nIF (-1.96<=resid<=1.96) THEN StudResidCount = 0;\n\nELSE IF (resid ne.) THEN StudResidCount = 1;\n\nTITLE 'Residual Count (StudResidCount = 0 should be >=95%)';\n\nRUN;\n\nPROC FREQ; TABLE StudResidCount; RUN;\n\nPROC LOESS DATA = voilib.residuals;\n\nODS OUTPUT OutputStatistics=voilib.smresid;\n\nMODEL resid = epoch / smooth = 0.4;\n\nBY group;\n\nRUN;\n\nSYMBOL1 i=none c=green v=dot;\n\nSYMBOL 2 i=none c=red v=dot;\n\nSYMBOL 3 i=spline c=green v=none l=2 w=5;\n\nSYMBOL 4 i=spline c=red v=none l=2 w=5;\n\nLEGEND1 label=('Residuals');\n\nLEGEND2 label=('Loess Curve');\n\naxis1 order=(-10 to 10 by 2);\n\nPROC GPLOT data=voilib.smresid;\n\nPLOT DepVar*epoch=group / vaxis = axis1 legend = LEGEND1;\n\nPLOT2 Pred*epoch=group / vaxis = axis1 legend = LEGEND2;\n\nTITLE 'Residuals by Time (should be no trend)';\n\nRUN;\n\n* plotting observed vs predicted *;\n\nSYMBOL1 I=NONE C=GREEN V=TRIANGLE;\n\nSYMBOL2 I=NONE C=RED V=HASH;\n\nSYMBOL3 I=JOIN C=GREEN V=TRIANGLE L=2 W=4;\n\nSYMBOL4 I=JOIN C=RED V=HASH L=2 W=4;\n\nLEGEND1 LABEL=('Observed');\n\nLEGEND2 LABEL=('Predicted');\n\nPROC GPLOT DATA=voilib.residuals;\n\nPLOT adjmean*epoch=group / LEGEND = LEGEND1; /* VAXIS = -7 to 7 ; * To set axis limits if different for each plot; */\n\nPLOT2 pred*epoch=group / LEGEND = LEGEND2; * VAXIS = -7 to 7 ;\n\nFORMAT adjmean 8.1;\n\nTITLE 'Predicted vs Observed';\n\nRUN;\n\nThe quantitative tests of normality in the proc univariate results give likelihoods of the each variable being normally distributed; for multivariable normality, all variables should be normally distributed, although this assumption is usually violated to some degree. The output of the UNIVARIATE procedure includes histograms of residuals (for assessing whether shape of distribution is approximately normal) and side-by-side box plots of residual distribution by group (Figure 3 and Figure 4).\n\nFurther possible tests include influence diagnostics, to determine whether particular subjects have undue effect on the model. If undue bias due to individual subjects is suspected, the approach we suggest is to use the influence measures to investigate whether one or more subjects are consistently and substantially discrepant from the others. In SAS these tests may be run by adding the INFLUENCE statement [e.g., after in PROC MIXED after “RESIDUAL” add “INFLUENCE (EFFECT=subject ITER=5)” and ODS SELECT INFLUENCE]; note that this requires SAS version 9.1 or greater. This test produces the restricted likelihood distance, PRESS statistic and Cook’s distance (known as Cook’s D) for each subject. If a discrepant subject is found, an attempt is made to determine whether there is some reason for this. If a reason can be found, then they might legitimately be excluded. A “sensitivity” analysis may be run (a comparison analysis including and excluding that patient) to see if there is a material difference in the conclusions or results of interest. The homogeneity of covariance matrices can also be assessed in a similar manner using the ESTIMATES option in the INFLUENCE statement. Although these influence diagnostics may be useful in some cases, they have less relevance in fMRI analyses due to the controlled nature of typical experiments and are presented here as optional.\n\nIf the residuals are non-normal but symmetrical around 0 then the model is probably robust. However, if the residuals are both non-normal and non-symmetrical (i.e., skewed) then estimate biases may appear. For a large number of patients or observations, results are typically robust to moderate amount of skew, but for smaller numbers the model is likely to be robust for no more than a small amount of skew. If the data are seriously skewed, the typical options include making transformations (log, power transformations, etc.), identifying and omitting unusual patients (see above), using another more appropriate analytical technique, introduce covariates into the model, or not analyze the data. For fMRI studies, omitting unusual patients and adding covariates are the most likely approaches. As mentioned earlier, if the findings from the model are considered of value, then we suggests researchers report them along with the diagnostics test results so that the reader can assess for themselves the validity of the results.\n\nRMANOVA has additional limitations which may be less applicable to fMRI data, but are detailed below for completeness. The constraints of the assumptions result in weaknesses of the approach, including:\n\nAlthough not usually relevant to fMRI studies, cases with missing data will be entirely deleted from the analysis, causing both conceptual and analytical difficulties. Imputation methods can be employed to circumvent this issue.\n\nTests of within-subjects effects assume sphericity, but this was not directly evaluated. The data can be tested for sphericity, but this evaluation is complex within the proc mixed framework, and is not addressed here. In theory, Mauchly’s test of spherericity can be used, for example, when any within-subjects factor has three or more trials. (If the within-subject factor fails to meet the assumption of sphericity, then either the multivariate approach can be used or the univariate results can be adjusted using a correction factor, e.g., the Huynh-Feldt Epsilon correction method [the Greenhouse-Geisser method has been shown to be too conservative]).\n\nThe choice of the within-subject correlation matrix form is dependent upon sphericity assumptions, and we suggest a compound symmetric matrix (“type = cs” option in proc mixed) to allow for different variance patterns across groups. If different variance patterns across time are suspected, then heterogeneous compound symmetry could be used (“type = csh” option). However, this process is extremely computationally intensive, and in tests we performed, made no notable difference to the fit of the model.\n\nFinally, the SAS library voilib created with the above SAS code contains numerous tables of data including predicted values, observations, model fit statistics, and residuals. These tables can be examined further either within SAS or by exporting them to a graph package, such as Excel.\n\n\nImplementation and results\n\nWe used fMRI data collected from two obstructive sleep apnea and two healthy control subjects as part of a pilot project; data are available online9. The data were chosen to illustrate the RMANOVA methodology, and do not constitute a sample that is sufficient to test scientific hypotheses. The Institutional Review Board of UCLA approved this study (IRB# 10-001012), which was in compliance with the Declaration of Helsinki; participants provided informed, written consent after the nature of the procedures was explained. Following a baseline scanning period, subjects performed a series of four respiratory challenges (30 s maximal inspiratory apnea), which were expected to elicit abnormal physiological responses in the patient group, based on earlier demonstrations of impaired neural responses to physiologic challenges10–13. A standard fMRI whole-brain protocol with repetition time of 2.5 s was implemented on a 3T Siemens Trio MRI scanner, and high-resolution T1-weighted anatomical scans were also collected (voxel size = 0.9 × 0.9 × 1 mm). The fMRI data were preprocessed using SPM routines, including realignment, spatial normalization, and smoothing. Smoothed images were intensity normalized to minimize global effects. VOI derived from the “AAL” toolbox in SPM were used to extract timetrends from the processed data14, and eight were selected to illustrate a variety of patterns highlighted by the approach. For each VOI, the time-series were analyzed using RMANOVA across the groups of two subjects, with challenges combined. Accompanying SAS and Excel files for each VOI are included online with this publication (Data availability).\n\nPreprocessing was performed using MATLAB 7 (The Mathworks, Inc., Natwick, MA, USA), SPM (http://www.fil.ion.ucl.ac.uk/spm/) and a custom detrending routine6. VOI were drawn using MRIcroN software15, and RMANOVA was implemented in SAS 9.47.\n\nA set of results is presented in Figure 6, illustrating a variety of statistical responses for eight VOI within the same fMRI dataset. This section describes the analytic steps needed to arrive at these results.\n\nVOI are from the Automatic Anatomical Labelling toolbox extension to SPM. Time-points of between and within-group significant responses relative to baseline period are indicated by color-coded “*” symbols. The graphs are in percent signal change relative to baseline, and each trace is a group/challenge average (± between-subject SE).\n\nResidual tests of the RMANOVA assumptions. The tests described in Appendix C were performed. The predicted responses were similar to the observed responses. Distributional tests of the Studentized residuals tests determined that the data were not precisely normally-distributed, as determined by the statistical tests with the PROC UNIVARIATE procedure. However, over 95% of the residuals were within ± 1.96, and visual inspection of the plots of residuals and predicted versus observed values showed only minor skew, and therefore the data were considered to be adequately modeled by the RMANOVA.\n\nPatterns highlighted by RMANOVA. All significant effects were BOLD signal responses. To infer neuronal responses, these patterns should be deconvolved with an HRF; this approach was not applied here.\n\nThe Tukey-Fisher criterion for multiple comparisons was applied, and the highlighted VOI did show significant overall effects (Figure 6A–H).\n\nThe three VOI in Figure 6A–C would all be highlighted as having significant group differences in response using conventional SPM analyses. However, RMANOVA is one way to illustrate the manner in which the groups differ. Figure 6A illustrates a typical fMRI pattern of difference, with one group showing increased signal throughout the challenge, and the other group essentially no change. Figure 6B illustrates an opposite response between the groups, with on increasing and the other decreasing during the challenge period. Figure 6C shows a decrease in one group with no change in the other.\n\nFigure 6D illustrates a simple case of a similar change in both groups; thus the within-group timepoints of difference are highlighted during the challenge period, corresponding to an equivalent decrease in OSA and control subjects. Figure 6E shows both groups changing in the same direction, but with a greater magnitude of response in the OSA group, thus leading to significant between-group differences across multiple time-points. Figure 6E also illustrates a pattern of changes that lasts over 30 seconds into the recovery period.\n\nTransient differences are shown in Figure 6F–H. Figure 6F shows a pattern of initial increased response in the OSA group and sustained decreased response in the control group. However, unlike previous examples, the trends in the OSA group change 10–15 seconds into the challenge from an increase to a decrease below baseline. Figure 6G shows a an opposite pattern in control subjects (initial decrease, latter increase), and in OSA two large peaks are identified as differing significantly both between group and within group relative to baseline. Finally, Figure 6H shows a switch between greater OSA signal at the start of the challenge followed by a lower signal during the second half of the task period. These examples illustrate the capability of RMANOVA to detect the timing of within and between group differences in fMRI data.\n\n\nDiscussion and conclusions\n\nThe results illustrate two advantages of RMANOVA: firstly, no prior model of expected response of either signal intensity or timing is assumed, and for the type of challenge shown here, a variety of significant response patterns was detected by RMANOVA. Secondly, once a significant effect is found, RMANOVA provides an objective, statistically rigorous assessment of the time when responses or differences occurred, and the precise response pattern. A group difference highlighted by a traditional SPM analysis does not differentiate between a group increase vs. no change, a group increase vs. a decrease, no change vs. a decrease, or a larger increase vs. a smaller increase.\n\nThe procedure allows analysis of multiple subjects within multiple groups. The within-group and between-group are all performed with one analysis, allowing the contributions of subject and group factors to be accounted for in the final results.\n\nNote that if all expected responses to the fMRI task are “on/off” or boxcar in nature, then the advantages of RMANOVA compared with cluster analysis are diminished.\n\nAnother advantage lies in processing timetrends from VOI drawn in different locations across different subjects. Spatial normalization is only accurate to within several millimeters16, and therefore small brain structures, such as those in the brainstem cannot be accurately depicted on VOI across multiple spatially normalized scans. By allowing a group analysis of timetrends from varying sites, RMANOVA can be more accurate compared with conventional fMRI analyses in depicting group responses within a particular structure. For regions such as the dorsal medulla, VOI analysis using RMANOVA is particularly helpful for determining group patterns.\n\nTwo disadvantages of RMANOVA relate to defining the VOI. The procedure requires manual definition of the VOI, perhaps for each subject, or at least for each structure of interest. A second disadvantage is that no whole-brain search is performed, and only a priori defined areas are studied. For this reason, we believe that a combination of a traditional SPM analysis with RMANOVA allows for a complete and powerful approach to analysis of fMRI data where the time-course of responses is of interest.\n\nThe approach does not consider or allow for variations in the shape of the HRF. A standard HRF may be used to deconvolve each time series, resulting in an inferred neural response time series, or a different HRF may be used for different VOI (to account for the spatial variation in HRF shape).\n\nTemporal autocorrelation between repeated measures is not accounted for, which likely leads to some loss of power. Nevertheless, we have found the technique highlights many patterns of interest, and thus this limitation does not negate the usefulness of the technique.\n\nProcedural disadvantages with the proposed method include the requirement for detrending images prior to analysis, the extraction of the VOI time-trends from the images, and in the example presented here, the use of SAS. The former requires computation routines which are typically not included in fMRI analysis software, and the latter requires routines for exporting the data, availability of software, and some familiarity with this software.\n\nThere are a number of assumptions underlying the RMANOVA method, which are fully described above along with their associated limitations. As with any statistical model, RMANOVA should be used with problems that match those assumptions.\n\nAnalyzing timetrends from VOI’s using RMANOVA can highlight temporal patterns of response to a challenge not readily apparent using conventional model-based approaches. Our research group has used this method extensively to assess fMRI responses to physiological challenges with complex responses over periods of tens of seconds to minutes. RMANOVA allows insight into the precise timing of changes from baseline, and response differences between groups. For complex paradigms, the technique can be a useful addition to conventional whole-brain model-based approaches.\n\n\nData and software availability\n\nF1000Research: Dataset 1. Processed data, as well as the SAS code for running each VOI analysis, 10.5256/f1000research.8252.d11747917\n\nHarvard Dataverse: Macey, Paul. 2016. Pilot fMRI of breath-hold in OSA, http://dx.doi.org/10.7910/DVN/EZUMI99",
"appendix": "Author contributions\n\n\n\nPMM and RMH conceived the study; PMM and PJS designed the methodology; PMM, KEM and RMH collected the test data and performed analysis and evaluation; all authors contributed to the manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nSupported by the National Institutes of Health National Institute of Nursing Research NR 013693.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nMacey KE, Macey PM, Woo MA, et al.: fMRI signal changes in response to forced expiratory loading in congenital central hypoventilation syndrome. J Appl Physiol (1985). 2004; 97(5): 1897–907. PubMed Abstract | Publisher Full Text\n\nOgren JA, Macey PM, Kumar R, et al.: Central autonomic regulation in congenital central hypoventilation syndrome. Neuroscience. 2010; 167(4): 1249–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMacey PM, Wu P, Kumar R, et al.: Differential responses of the insular cortex gyri to autonomic challenges. Auton Neurosci. 2012; 168(1–2): 72–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPenny W, Friston K: Mixtures of general linear models for functional neuroimaging. IEEE Trans Med Imaging. 2003; 22(4): 504–14. PubMed Abstract | Publisher Full Text\n\nFriston KJ, Holmes AP, Poline JB, et al.: Analysis of fMRI time-series revisited. Neuroimage. 1995; 2(1): 45–53. PubMed Abstract | Publisher Full Text\n\nMacey PM, Macey KE, Kumar R, et al.: A method for removal of global effects from fMRI time series. Neuroimage. 2004; 22(1): 360–6. PubMed Abstract | Publisher Full Text\n\nLittell RC, Milliken GA, Stroup WW, et al.: SAS System for Mixed Models. Cary, NC: SAS Institute Inc.; 1996. Reference Source\n\nMacey PM, Valderama C, Kim AH, et al.: Temporal trends of cardiac and respiratory responses to ventilatory challenges in congenital central hypoventilation syndrome. Pediatr Res. 2004; 55(6): 953–9. PubMed Abstract | Publisher Full Text\n\nPilot fMRI of breath-hold in OSA. Harvard Dataverse. 2016. Publisher Full Text\n\nMacey KE, Macey PM, Woo MA, et al.: Inspiratory loading elicits aberrant fMRI signal changes in obstructive sleep apnea. Respir Physiol Neurobiol. 2006; 151(1): 44–60. PubMed Abstract | Publisher Full Text\n\nMacey PM, Macey KE, Henderson LA, et al.: Functional magnetic resonance imaging responses to expiratory loading in obstructive sleep apnea. Respir Physiol Neurobiol. 2003; 138(2–3): 275–90. PubMed Abstract | Publisher Full Text\n\nHenderson LA, Woo MA, Macey PM, et al.: Neural responses during Valsalva maneuvers in obstructive sleep apnea syndrome. J Appl Physiol (1985). 2003; 94(3): 1063–74. PubMed Abstract | Publisher Full Text\n\nHarper RM, Macey PM, Henderson LA, et al.: fMRI responses to cold pressor challenges in control and obstructive sleep apnea subjects. J Appl Physiol (1985). 2003; 94(4): 1583–95. PubMed Abstract | Publisher Full Text\n\nTzourio-Mazoyer N, Landeau B, Papathanassiou D, et al.: Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage. 2002; 15(1): 273–89. PubMed Abstract | Publisher Full Text\n\nRorden C, Karnath HO, Bonilha L: Improving lesion-symptom mapping. J Cogn Neurosci. 2007; 19(7): 1081–8. PubMed Abstract | Publisher Full Text\n\nKrishnan S, Slavin MJ, Tran TT, et al.: Accuracy of spatial normalization of the hippocampus: implications for fMRI research in memory disorders. Neuroimage. 2006; 31(2): 560–71. PubMed Abstract | Publisher Full Text\n\nMacey P, Schluter P, Macey K, et al.: Dataset 1 in: Detecting variable responses within fMRI time-series of volumes-of-interest using repeated measures ANOVA. F1000Research. 2016. Data Source"
}
|
[
{
"id": "13288",
"date": "26 Apr 2016",
"name": "Joke Durnez",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper aims to present a different approach to analyzing fMRI time trends. The idea is to average the signal overall trials, over subjects (similar to ERP-analysis) and look at the differences between groups and conditions and especially its interaction with time since onset. Because the trials are clustered within subjects, the data points are modeled using a mixed model. The result is a figure with the average time course per group/condition, with a statistical test comparing the BOLD signal at each time point.The main difference between this model and the model in the standard (frequentist, parametric mixed model) procedure is the specific within-subjects design matrix. The predictor over time in a standard analysis would be:[ 0 0 0 1 1 1 0 0 0 1 1 1] o HRFThe predictor over time with the presented approach would be a categorical variable:[ 0 0 0 1 2 3 0 0 0 1 2 3 ]The model tested is then BOLD = GROUP + TIME + GROUP*TIME. Important: TIME is modeled as a categorical variable, not continuous. Modeling time as a categorical folder allows to test for each time point separately as a post-hoc test. The results of interest are the post-hoc tests of the interaction effect between group and time.SOME COMMENTSThe focus of the paper is on the post-hoc tests. These are definitely interesting, but the post-hoc tests should only be performed if the interaction group*time is significant. It would be interesting to also have some more detail on the main model (e.g. main effect of time, what does it mean? what is the interaction?) In my opinion the chunks of code make it hard to read through the paper. It would be easier if the methods section only describes the different choices/tests/parameters and the code is grouped in an appendix. Especially code that is just for the purpose of making variables or figures should not be in the main manuscript. In general, I think SAS is a bad choice. Even though it has definitely one of the best mixed model implementations, it’s not a typical language at all in neuroimaging. If you want people to use this procedure, consider having a script in R, matlab or python. I have tried to reproduce the code with the data attached. However, the datasets have different names and paths. It’s hard to derive what is what… I very much liked the diagnostics, something that is most often lacking in neuroimaging studies. I think the part about missing data could be removed. This problem is non-existing in fMRI. In the results section: “all significant effects were BOLD signal responses”. Poor choice of words. The outcome variable is the BOLD signal response, not the effect. “the highlighted VOI did show significant overall effects”: significant overall effects of group? of time? the interaction? all of them? The interpretation given to the time series is the interpretation of a pattern. However, these tests are not tests for patterns. For example, the interpretation of the interaction on fig 6E is not significant on all time points but is interpreted as such. This door is wide open for over interpreting non-significant results. This is a problem with these point wise comparisons. There are ways to take away the temporal autocorrelation (whitening), which is currently used in fMRI software. You should mention this when talking about the problem of temporal correlation.",
"responses": [
{
"c_id": "2050",
"date": "08 Jul 2016",
"name": "Paul Macey",
"role": "Author Response",
"response": "[Initial comments both reviewers] Thanks to the reviewers for their thoughtful comments. We made extensive revisions to the paper, which should make this method more accessible. The major changes including adding R code for implementing RMANOVA and diagnostic tests, simplifying some of the SAS commands and formatting requirements, and clarifying the nature of time series for which the method is applicable. We did investigate implementing the method in MATLAB, but found that approach to be cumbersome and limited. While we did not test the method in SPSS, our understanding is that such implementation is similar to SAS, and relatively straight forward. In R, RMANOVA is non-trivial, and we settled on a mixed model approach that is equivalent to the one presented in SAS, but with more limited options should the user wish to try alternative covariance options. Furthermore, in R there are many options for diagnostic plots and calculations, and we selected one series of commands and packages that implement all SAS outputs other than studentized residuals. In proposing a simplified data format, we decided to expand the scope of the method to include physiologic signals other than fMRI, such as heart rate, pulse oximetry, and other continuously acquired signals. We have applied our RMANOVA to these other signals in several publications. We therefore changed the title of the paper to reflect this broader application. We updated the “dataverse” repository to include example R and SAS code, and a formatted data file. Specific responses to Joke Durnez's comments This paper aims to present a different approach to analyzing fMRI time trends. The idea is to average the signal overall trials, over subjects (similar to ERP-analysis) and look at the differences between groups and conditions and especially its interaction with time since onset. Because the trials are clustered within subjects, the data points are modeled using a mixed model. The result is a figure with the average time course per group/condition, with a statistical test comparing the BOLD signal at each time point. The main difference between this model and the model in the standard (frequentist, parametric mixed model) procedure is the specific within-subjects design matrix. The predictor over time in a standard analysis would be: [ 0 0 0 1 1 1 0 0 0 1 1 1] o HRF The predictor over time with the presented approach would be a categorical variable: [ 0 0 0 1 2 3 0 0 0 1 2 3 ] The model tested is then BOLD = GROUP + TIME + GROUP*TIME. Important: TIME is modeled as a categorical variable, not continuous. Modeling time as a categorical folder allows to test for each time point separately as a post-hoc test. The results of interest are the post-hoc tests of the interaction effect between group and time. We appreciate this simplified description of the technique; we have included a synthesis at the start of the Introduction. SOME COMMENTS The focus of the paper is on the post-hoc tests. These are definitely interesting, but the post-hoc tests should only be performed if the interaction group*time is significant. It would be interesting to also have some more detail on the main model (e.g. main effect of time, what does it mean? what is the interaction?) The sequence is to test 1) global fit of the model; if significant then test 2) variable and interaction level fit; and if significant test 3) post hoc at individual time points. This sequence was not clear in the original version. We agree that pot hoc assessments at time points are only valid for significant model-level effects of time (within group) or group*time (between group), and we have emphasized this point in the revised paper. In my opinion the chunks of code make it hard to read through the paper. It would be easier if the methods section only describes the different choices/tests/parameters and the code is grouped in an appendix. Especially code that is just for the purpose of making variables or figures should not be in the main manuscript. We have simplified the code and proposed input data format, which eliminates some of the more cumbersome and less relevant code. We did originally have the code in appendices, but this format did not fit the requirements of this journal. In the revised paper, we added comments and restructured some sections to improve readability. In general, I think SAS is a bad choice. Even though it has definitely one of the best mixed model implementations, it’s not a typical language at all in neuroimaging. If you want people to use this procedure, consider having a script in R, matlab or python. We now provide an implementation in R, since this software is free and well supported in the statistical community. We did evaluate RMANOVA in MATLAB, but found no reasonable approach to replication the implementation we propose. While SAS is an environment principally familiar to statisticians, the implementation of the procedure and diagnostic tests is simpler and more comprehensive than R. For that reason, we have left the SAS code in the paper. I have tried to reproduce the code with the data attached. However, the datasets have different names and paths. It’s hard to derive what is what… We simplified the input data format to be a text file with three or four columns. The original approach of importing from Excel was cumbersome, but with the more logical format users should find it easier to run the SAS or R commands. I very much liked the diagnostics, something that is most often lacking in neuroimaging studies. We appreciate the comment. We have added explanations, as noted in response to Matthew Brett. I think the part about missing data could be removed. This problem is non-existing in fMRI. As noted in response to Matthew Brett , we have expanded the paper to explain how this RMANOVA approach applies to non-fMRI physiologic data, where missing values are possible. We have clarified the limitation. In the results section: “all significant effects were BOLD signal responses”. Poor choice of words. The outcome variable is the BOLD signal response, not the effect. We revised this introductory paragraph. “the highlighted VOI did show significant overall effects”: significant overall effects of group? of time? the interaction? all of them? We revised the wording in this section. The interpretation given to the time series is the interpretation of a pattern. However, these tests are not tests for patterns. For example, the interpretation of the interaction on fig 6E is not significant on all time points but is interpreted as such. This door is wide open for over interpreting non-significant results. This is a problem with these point wise comparisons. The reviewer is correct to note RMANOVA is not identifying patterns, but rather effects at time points independent of each other. We have changed this terminology. We have emphasized that the time points will only be assessed post hoc if the corresponding independent variable is significant within the model. If the group×time interaction is not significant, the between-group time points will not be considered. If the time variable is not significant, the within-group time points will not be considered. There are ways to take away the temporal autocorrelation (whitening), which is currently used in fMRI software. You should mention this when talking about the problem of temporal correlation. We have expanded this paragraph and included a note about whitening."
}
]
},
{
"id": "13203",
"date": "04 May 2016",
"name": "Matthew Brett",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper is shows how to apply the SAS \"MIXED\" procedure to FMRI data from ROIs, presents some data from this application, and discusses some of the options and diagnostics that this command provides.The idea is to do what other packages would call an FIR analysis. You first decide on a time-bin width, typically one scan, but sometimes more. Call the time bin width B seconds. Then the scan data is labeled according to how many time bin widths it is after some event onset up until a cutoff of N time-bins.Usually N * B is equal to the length of the event plus 20 seconds or so to let the hemodynamic response drop back to baseline. Then the fitting procedure calculates adjusted means for each for the N bins, giving an evoked response time-course that can simultaneously adjust for other effects in the data, fit in the regression model.In what follows, I confess I know SPM much better than the other packages, so I'll mainly compare to SPM.SPM is rather general, but it does have this capability. The standard procedure would be to fit an \"FIR\" model to each subject, giving an estimated signal for each time bin in beta images, with one beta image for each bin. You then take these time-bin beta images into your second-level model, where you can adjust for variance covariance effects such as correlation between time-bin values within subjects, different within-subject variance, and so on.As you say, using these methods in SPM can be moderately hard, but I'd argue the complexity is comparable to switching languages from MATLAB to SAS and using SAS to calculate things like adjusted means, residuals and so on.How would you expect the results of using the SPM interface (I guess within marsbar) to compare to the results of using the SAS methods you have here.I was expecting to see you using either MATLAB or R to do this; MATLAB because SPM already requires MATLAB, and it's a general programming language, R because it has a variety of procedures for fitting repeated measures models (see for example http://www.r-statistics.com/2010/04/repeated-measures-anova-with-r-tutorials/), is free, open-source, and, in my experience, is more widely used than SAS in researchcommunities, including neuroimaging.I'm afraid I haven't used SAS since the 1980s, so it's hard for me to comment on the details of the SAS commands.I found myself skip-reading the actual SAS commands, I don't know how useful they will be for someone not experienced with SAS. Maybe in-line comments would help?At various points you recommend options for diagnostics and outlier detection. It would help to have more detail as to why you recommend those options over others.Description of procedure, first paragraph \"Proportional scaling\" is a technical term (at least in SPM-land) for dividing each scan voxel value by the calculated in-brain mean across voxels for the same scan. That is now very rarely used SPM, and was never widely used for FMRI. The standard SPM scaling involves multiplying each scan in a run by the same scaling factor, that brings the mean of in-brain mean voxel values to 100.Assumptions, diagnostic checks, and limitations, first paragraph:Could you expand on \"Violations of independence produce a non-normal distribution of the residuals\"?Options if diagnostic tests fail, first paragraph \"but for smaller numbers the model is likely to be robust for no more than a small amount of skew.\". Could you rephrase to something like \"For small numbers, even a small amount of skew will make it unlikely that the model is robust.\". For some reason I found your original sentence a bit difficult to parse.Disadvantages of RMANOVA, third paragraph:You mention that \"Temporal autocorrelation between repeated measures is not accounted for\". Could you elaborate? I guess you mean that repeated measures variance / covariance adjustments in MIXED are for the estimates across time bins, not for the autocorrelation between scans. In that case, are youexpecting something like SPM to do better, given that it does adjust for autocorrelation at the first level?",
"responses": [
{
"c_id": "2049",
"date": "08 Jul 2016",
"name": "Paul Macey",
"role": "Author Response",
"response": "[Initial comments repeated for both reviewers] Thanks to the reviewers for their thoughtful comments. We made extensive revisions to the paper, which should make this method more accessible. The major changes including adding R code for implementing RMANOVA and diagnostic tests, simplifying some of the SAS commands and formatting requirements, and clarifying the nature of time series for which the method is applicable. We did investigate implementing the method in MATLAB, but found that approach to be cumbersome and limited. While we did not test the method in SPSS, our understanding is that such implementation is similar to SAS, and relatively straight forward. In R, RMANOVA is non-trivial, and we settled on a mixed model approach that is equivalent to the one presented in SAS, but with more limited options should the user wish to try alternative covariance options. Furthermore, in R there are many options for diagnostic plots and calculations, and we selected one series of commands and packages that implement all SAS outputs other than studentized residuals. In proposing a simplified data format, we decided to expand the scope of the method to include physiologic signals other than fMRI, such as heart rate, pulse oximetry, and other continuously acquired signals. We have applied our RMANOVA to these other signals in several publications. We therefore changed the title of the paper to reflect this broader application. We updated the “dataverse” repository to include example R and SAS code, and a formatted data file. Specific responses to Matthew Brett's comments The reviewer’s comments are helpful in clarifying how this method fits with other available approaches. This paper shows how to apply the SAS \"MIXED\" procedure to FMRI data from ROIs, presents some data from this application, and discusses some of the options and diagnostics that this command provides. The idea is to do what other packages would call an FIR analysis. You first decide on a time-bin width, typically one scan, but sometimes more. Call the time bin width B seconds. Then the scan data is labeled according to how many time bin widths it is after some event onset up until a cutoff of N time-bins. Usually N * B is equal to the length of the event plus 20 seconds or so to let the hemodynamic response drop back to baseline. Then the fitting procedure calculates adjusted means for each for the N bins, giving an evoked response time-course that can simultaneously adjust for other effects in the data, fit in the regression model. In what follows, I confess I know SPM much better than the other packages, so I'll mainly compare to SPM. SPM is rather general, but it does have this capability. The standard procedure would be to fit an \"FIR\" model to each subject, giving an estimated signal for each time bin in beta images, with one beta image for each bin. You then take these time-bin beta images into your second-level model, where you can adjust for variance covariance effects such as correlation between time-bin values within subjects, different within-subject variance, and so on. As you say, using these methods in SPM can be moderately hard, but I'd argue the complexity is comparable to switching languages from MATLAB to SAS and using SAS to calculate things like adjusted means, residuals and so on. An SPM FIR analysis could replicate the SAS implementation of our timetrend RMANOVA approach, and we have used FIR in single-group fMRI studies of physiologic challenges. However, the results of a multi-group FIR quickly become complex since there is a voxel-level map of significant responses at each time point during the challenge and recovery. Performing between-group comparisons with FIR requires quite large and complex design matrices and contrasts. The marsbar toolbox allows extraction of the timetrend for those contrasts at pre-determined VOI, but to indicate timepoints of significant within or between group responses, the user would still need to identify these manually from the bin contrast maps (at least that is my understanding). We agree that SAS is a complex environment unfamiliar to most neuroimaging researchers. At the time we originally developed this approach, MATLAB and R did not have the necessary capabilities, but that has mostly changed. In our revised paper, we present an implementation in R. We evaluated MATLAB but found no reasonable way to implement the present type of RMANOVA. In practice, the method is automated, so users don’t need to work within R or SAS other than running the analyses. One advantage of SAS over R is the option to easily choose a variety of covariance matrices. This could be important for experiments where the variability changes over time. For example, the standard “compound symmetric” (cs) covariance matrix we propose assumes equal variability at all timepoints including before during and after a task, whereas, for example, the “compound symmetric heterogenous” (csh) covariance matrix allows for different variances (and hence standard error estimates) at each timepoint. Theoretically csh should allow for more accurate models, although we found no major differences for our physiologic experiments and in our example use cs due to the lower computation needs. The R implementation does not have the option to choose covariance matrices, but we used a random intercept for each subject, which is equivalent to cs. How would you expect the results of using the SPM interface (I guess within marsbar) to compare to the results of using the SAS methods you have here. The principal advantage of our technique is that it gives simple results that match the fMRI time-trend patterns. If a timetrend is higher or lower than the baseline at a timepoint, the method will indicate whether that increase was significant or not via a timepoint-specific p-value. If the timetrends for two groups diverge at a timepoint, the method will indicate whether that group difference is significant, again with a timepoint-specific p value. I was expecting to see you using either MATLAB or R to do this; MATLAB because SPM already requires MATLAB, and it's a general programming language, R because it has a variety of procedures for fitting repeated measures models (see for example http://www.r-statistics.com/2010/04/repeated-measures-anova-with-r-tutorials/), is free, open-source, and, in my experience, is more widely used than SAS in research communities, including neuroimaging. I'm afraid I haven't used SAS since the 1980s, so it's hard for me to comment on the details of the SAS commands. I found myself skip-reading the actual SAS commands, I don't know how useful they will be for someone not experienced with SAS. Maybe in-line comments would help? We have provided an R implementation, and added comments. At various points you recommend options for diagnostics and outlier detection. It would help to have more detail as to why you recommend those options over others. We have added additional comments explaining the tests. Description of procedure, first paragraph \"Proportional scaling\" is a technical term (at least in SPM-land) for dividing each scan voxel value by the calculated in-brain mean across voxels for the same scan. That is now very rarely used SPM, and was never widely used for FMRI. The standard SPM scaling involves multiplying each scan in a run by the same scaling factor that brings the mean of in-brain mean voxel values to 100. We have removed this comment. Assumptions, diagnostic checks, and limitations, first paragraph: Could you expand on \"Violations of independence produce a non-normal distribution of the residuals\"? We added some further explanation of independence violations. Options if diagnostic tests fail, first paragraph \"but for smaller numbers the model is likely to be robust for no more than a small amount of skew.\". Could you rephrase to something like \"For small numbers, even a small amount of skew will make it unlikely that the model is robust.\". For some reason I found your original sentence a bit difficult to parse. We agree the original wording was confusing, and we have reworded according to this suggestion. Disadvantages of RMANOVA, third paragraph: You mention that \"Temporal autocorrelation between repeated measures is not accounted for\". Could you elaborate? I guess you mean that repeated measures variance / covariance adjustments in MIXED are for the estimates across time bins, not for the autocorrelation between scans. In that case, are you expecting something like SPM to do better, given that it does adjust for autocorrelation at the first level? We give more detail in this paragraph. Specifically, we note that if there is a high correlation between adjacent time points (that is, there is only slow variation), this limitation will be more prominent. If the response are not expected to change much over time, SPM would be a better approach. We try to capture this aspect with “variable responses” in the paper title."
}
]
}
] | 1
|
https://f1000research.com/articles/5-563
|
https://f1000research.com/articles/5-885/v1
|
13 May 16
|
{
"type": "Research Article",
"title": "Priority regions for research on dryland cereals and legumes",
"authors": [
"Glenn Hyman",
"Elizabeth Barona",
"Chandrashekhar Biradar",
"Edward Guevara",
"John Dixon",
"Steve Beebe",
"Silvia Elena Castano",
"Tunrayo Alabi",
"Murali Krishna Gumma",
"Shoba Sivasankar",
"Ovidio Rivera",
"Herlin Espinosa",
"Jorge Cardona",
"Elizabeth Barona",
"Chandrashekhar Biradar",
"Edward Guevara",
"John Dixon",
"Steve Beebe",
"Silvia Elena Castano",
"Tunrayo Alabi",
"Murali Krishna Gumma",
"Shoba Sivasankar",
"Ovidio Rivera",
"Herlin Espinosa",
"Jorge Cardona"
],
"abstract": "Dryland cereals and legumes are important crops in farming systems across the world. Yet they are frequently neglected among the priorities for international agricultural research and development, often due to lack of information on their magnitude and extent. Given what we know about the global distribution of dryland cereals and legumes, what regions should be high priority for research and development to improve livelihoods and food security? This research evaluated the geographic dimensions of these crops and the farming systems where they are found worldwide. The study employed geographic information science and data to assess the key farming systems and regions for these crops. Dryland cereal and legume crops should be given high priority in 18 farming systems worldwide, where their cultivated area comprises more than 160 million ha. These regions include the dryer areas of South Asia, West and East Africa, the Middle East and North Africa, Central America and other parts of Asia. These regions are prone to drought and heat stress, have limiting soil constraints, make up half of the global population and account for 60 percent of the global poor and malnourished. The dryland cereal and legume crops and farming systems merit more research and development attention to improve productivity and address development problems. This project developed an open access dataset and information resource that provides the basis for future analysis of the geographic dimensions of dryland cereals and legumes.",
"keywords": [
"Grain legumes",
"dryland cereals",
"farming systems",
"priority setting",
"geographic priorities"
],
"content": "Introduction\n\nInternational agricultural research and development programs usually consider the geographic dimensions of crop improvement and farming systems in their efforts to prioritize activities (Hyman et al., 2013). Where is the crop grown globally and what are the key obstacles to crop production? Does the research apply to places where benefits can reach a large number of people? How can resources be allocated to achieve efficiencies? But answering these questions requires integration of socioeconomic and biophysical data and the integration of wide-ranging data and information resources. Often data exists for national jurisdictions, but it needs to be evaluated by agroecology or farming system. The task is even more difficult for crops such as dryland cereals and grain legumes because – when compared to the major staple crops – these crops are often embedded in complex crop-livestock systems and less information is available. The Dryland Cereals and Legumes Agri-Food System research program of the CGIAR (the program is hereafter referred to as DCL) requested an analysis of the principal commodities of their proposed program and the farming systems in which they are found. The 12 priority crops of these dryland systems are chickpea, common bean, cowpea, faba bean, groundnut, lentil, pigeon pea, soybean, barley, pearl millet, small millet and sorghum (DCL, 2015). The research presented in this paper shows the development of spatial and statistical data intended to support geographic priority setting for the global DCL research program. In order to develop the analysis, this research builds on a global classification of farming systems, on maps of the spatial distribution of all 12 DCL crop commodities, on socioeconomic data on population, poverty, malnutrition, on market access, and on soil and climatic data. The analysis identifies where these crops occur in the context of constraints and opportunities for their development. How can DCL technologies be geographically targeted for achieving the objective of reducing poverty and malnutrition in dryland systems? The present analysis is based on a diverse array of geographic information, and includes new assessments of poverty, drought, heat and other information related to crop improvement and management. In this way, the study examines the spatial extents of key constraints to DCL crop production, using the most recent spatial data available. The analysis and resulting database provides the first global farming systems information resource for specifically evaluating priorities for DCL crop improvement and management. Tabular data from this analysis is open access and has been published in a data repository (Barona et al., 2016a). The geospatial data used this study can be accessed through a new online digital atlas for dryland cereals and grain legumes (see http://www.eatlasdcl.cgiar.org/).\n\n\nMethodology\n\nThis study builds on previous work (including Dixon et al., 2001 and Hyman et al., 2008), but with a focus on the 12 principal commodities and farming systems of DCL. The main framework for the study is John Dixon’s farming systems framework, a global delineation and resulting map of the major farming systems of the developing world (Dixon et al., 2001). Dixon’s schema is built on 15 biophysical and socioeconomic spatial data layers available in the year 2000 and consultations with hundreds of regional and global agricultural experts using a modified Delhi technique. In a participatory process, 72 global farming systems in six developing regions were geographically delineated and their characteristics described. The present study uses a subset of 63 of those farming systems together with new spatial data on biophysical and socioeconomic conditions to characterize the extents of DCL commodities. Using spatial overlay, biophysical and socioeconomic information are organized according to the 63 Dixon farming systems.\n\nA key advantage of this research was that instead of analyzing crop information by country, subnational estimates of crop distribution are generated based on pixel level data (Hyman et al., 2008). Then, using spatial overlay, we organize that data by country (250 in total), by farming system (63 types) and by combinations of countries and farming systems (544 combinations). Other data is also organized according to farming system and country – including information on drought, temperature dynamics with climate change, soil conditions, population and poverty. Readers should consult our previous publication and available data for further details on the data and methodology (Barona et al., 2016a and Barona et al., 2016b; Hyman et al., 2008; Hyman et al., 2015).\n\nSpatial information on biophysical and socioeconomic conditions was acquired, with the objective of obtaining the most recent and spatially detailed information related to dryland cereals and legume and the agricultural systems where they are found. The present study upgrades our previous work because we are using data that was not available before, especially the 2005 spatial distribution of crop area, production and yield (You & Wood, 2006; You et al., 2014a; You et al., 2014b). The previous study used crop distribution data from the year 2000, while the work we describe here uses 2005 crop distribution data. Our previous dataset only included 7 DCL commodities, in contrast to all 12 of the DCL commodities used here. These new data also benefited from improved spatial resolution and modeling procedures. This study used the most recent available data on global livestock and human population. The source of the year 2010 human population data was the gridded population of the world project (CIESIN, 2014). Livestock population was taken from the Gridded Livestock of the World (GLW) database at 5 km spatial resolution, with the year 2005 as the reference year (Robinson et al., 2014).\n\nSeveral datasets gave us information on abiotic constraints to crop production that are important for the DCL commodities. The dataset includes indicators of drought based on maps of drought probability and the “failed seasons” concept. By simulating rainfall for defined crop water requirements, the probability of a growing season failing to produce a successful harvest indicates drought risk for every pixel across the world (Hyman et al., 2008; Jones & Thornton, 2000; Jones et al., 2002). The drought probability is multiplied by total crop area to derive the potential drought impact index (PDII). Furthermore, the needed heat tolerance for DCL crops was indicated by estimates of expected temperature change between the current temperature and 2050 temperatures (Hijmans et al., 2005, Ramirez & Jarvis, 2008). These predicted changes are based on global circulation models (GCM) under the A1B scenario, assuming rapid economic growth with emissions peaking around 2050. The study used maps of soil constraints based on the fertility capability classification (Sanchez et al., 2003). These constraints included soil acidity, length of the dry season, waterlogging, low nutrient availability and salinity – all constraints identified by DCL crop experts as important obstacles to overcome (DCL, 2015). Finally, the length of the growing period indicates seasonal constraints on crops that may be relevant for the breeding objectives of DCL crops (Fischer, 2009).\n\nDetailed geographic information on population is not usually available until at least five years after the dates of censuses and surveys. Our analysis includes estimates of the total population for the year 2010, as well as total, rural and urban population for 2005 (CIESIN et al., 2005; CIESIN, 2014). We included 2005 population data in our analysis because the CIESIN (2014) population dataset does not yet include urban and rural data for 2010. The analysis draws on estimates of the number of people living on less than $1 and $2 per day for 10 km pixel areas, based on estimates derived from combined poverty maps and survey data for the entire world, with a base year of 2005. This global poverty data set is not available in the public domain, but interested users of poverty data should consult the HarvestChoice website to learn more about this and other poverty mapping initiatives (Stanley Wood, personal communication). Nutrition indicators include the absolute number and proportional numbers of children under five years old that are two standard deviations below the median of weight for age (underweight) and height for age (stunting), according to international standards (CIESIN, 2005; FAO, 2007).\n\nSpatial overlay was used to organize the data into spatial units according to farming system and combinations of farming systems and country. All spatial data was converted to the Robinson equal area projection at 10 km spatial resolution before processing commenced. We used the zonal statistics tools in ArcInfo Workstation 10.0 and ArcGIS Desktop 10.1 software. The analysis digitally overlays Dixon’s farming systems map and a global map of country boundaries on the socioeconomic and biophysical map data described above. The result of the overlay procedure is a set of database files (dBase format) organized by farming system region and combination of farming system region and country. The database files were then converted to 40 spreadsheet files in Excel format (Barona et al., 2016b). An additional analysis was made of the pixels where more than one DCL crop occurred within the pixel. For each crop, if the area value in the pixel was higher than the mean for all pixels of that crop, it was considered to be of a sufficient density to map these crop combinations. By selecting only those pixels above the mean, we excluded those areas that may have a small concentration of the crop. The creation of the tables was facilitated using scripts written in Arc Macro Language (AML) to facilitate updates as more recent or better data becomes available (scripts and data available from Barona et al., 2016b).\n\nThe study used a mix of criteria for determining priority regions for research and development in DCL farming systems. A modified “natural breaks” approach was our primary consideration in selecting priority farming system regions. Building on the determination of classes for choropleth mapping, the approach visually inspects the data to find where farming system regions group together according to their levels of DCL crop area (Smith, 1986). We also considered whether a single crop dominated a farming system region in areas with large farms and well-developed agro-industries. Some farming systems that have relatively small DCL crop areas could be included if they were similar to regions that have large areas, something that often occurs with farming systems in different regions of the world.\n\nAnother criteria was whether these farming systems overlapped with DCL program target countries, as established in the DCL pre-proposal (DCL, 2015). A key criteria for program target countries is that they fall within dryland regions, defined as areas with an aridity index between 0.03 and 0.65 (Zomer et al., 2008). A dryland regions map and a description of how it was made can be accessed on the online DCL Atlas. Beyond this consideration, the DCL program’s target country analysis considered crop area and production, people living in poverty, childhood malnutrition, land degradation and other considerations – all with national level data. The reference map in the online atlas contains the map of target countries of the DCL program.\n\nThe approach we describe above leaves open the possibility to select different criteria and to expand or contract the number of farming systems that would be targeted for a research and development program. The publication of replication data found in the online atlas and in the Dataverse repository enable future iterations of the analysis according to any adjustments that the program may want to make (Barona et al., 2016a).\n\n\nResults\n\nThe DCL crops are concentrated in 18 farming systems where more than 160 million ha of these crops are cultivated, where more than 60% of the world’s poor live and where the DCL commodity programs have selected target countries based on their fit with dryland systems (Figure 1; Table 1 & Table 7; DCL, 2015; DESA, 2009). We selected these farming systems if the farming system had at least six million ha of combined DCL crops. However, we excluded three Latin American and Caribbean (LAC) farming systems that met this threshold because these systems were overly dominated by soybean production in regions with typically large farms. These excluded systems were temperate mixed (Pampas), cereal-livestock (Campos) and extensive mixed (Cerrados_Llanos).\n\nPriority DCL farming systems were included from Latin America and the Middle East and North Africa (MENA) regions that did not meet the six million hectare threshold described above. The maize-beans farming system in Mesoamerica was added because it is very similar to maize mixed system in sub-Saharan Africa. The rainfed mixed and pastoral farming systems in the MENA region were added because they are similar to farming systems in sub-Saharan Africa (pastoral) and in South Asia (rainfed mixed). Two other farming systems – dry rainfed and highland mixed – are included on the basis of traditional importance in the dryland MENA region.\n\nThree farming systems in South Asia – rainfed mixed, rice-wheat and dry rainfed – make up about one third of the 162 million ha of DCL crops in the 18 priority farming systems (Table 1). The rainfed mixed system makes up 20% of the DCL crop area in these priority farming systems, accounting for more than 30 million ha of DCL crops. A second important region is Sub-Saharan Africa, where the cereal-root crop mixed system accounts for 21.3 million ha, the agro-pastoral millet sorghum system accounts for 18.6 million ha, the pastoral system accounts for 10.8 million ha and the maize mixed system has 7.6 million ha. In Eastern Europe and Central Asia more than 15 million ha are cultivated, with barley figuring prominently. In East Asia over 22 million ha are cultivated, with groundnut and soybean as the predominant crops. The overall DCL crop area in the Middle East and North Africa is the lowest among the 18 priority regions suggested above.\n\nIn some cases DCL crops make up a large proportion of the total cultivated area in these farming systems, but their overall area may be relatively small when they are found in systems with large areas of maize, wheat and rice (Table 2). Three cereals (barley, pearl millet and sorghum) and two legumes (soybean and cowpea) play key roles in several farming systems where they make up more than 10% of all cultivated crop area within the system. The legumes are multi-purpose, contributing soil fertility, family nutrition, fodder and cash sales. Seven relatively cooler farming systems have more than 13 percent of their crop area in barley – four of which are in the Middle East and North Africa and three in Eastern Europe and Central Asia region. Pearl millet is an important component in the higher temperature pastoral, agro-pastoral millet-sorghum and cereal-root crop mixed farming systems, making up 35%, 32% and 11% of the total cultivated area respectively. In the pastoral and agro-pastoral millet-sorghum systems pearl millet is a major food crop, whereas in the somewhat higher rainfall cereal-root crop mixed system maize, sorghum and cassava are the major food crops. In three African systems and one South Asian system, sorghum makes up more than 20% of the total cultivated area, although being pushed back by drought tolerant maize. In five of these 18 farming systems groundnuts make up between five and eight percent of the total cultivated area. The remaining crops – bean, chickpea, lentil, small millet and pigeon pea – have a smaller overall agricultural footprint.\n\nIn some cases DCL crops make up a large proportion of the total cultivated area in these farming systems, but their overall area may be relatively low when they are found in systems with large areas in maize, wheat and rice.\n\nYields vary across the 18 farming systems and by DCL commodity (Table 3). A very general pattern is that yields are lowest in sub-Saharan Africa. They are somewhat higher in South Asia and even more so in Eastern Europe and Central Asia. Finally they are highest in the East Asian countries. These differences are related to many different factors, including population density, access to agricultural services and markets, biotic and abiotic constraints, technology levels, management practices and others.\n\nLivestock is an important component of the DCL research program and all the farming systems where DCL crops are concentrated. The DCL crops are considered multi-purpose crops because they are used for many purposes including food, feed and fodder – as well as ecosystem services. Soybean and barley are perhaps the most important for livestock, with much of their production going towards animal fodder (Hartman et al., 2011; Newton et al., 2011). Sorghum and millet is also very important as feed and fodder in sub-Saharan Africa. Table 4 shows the estimated 2005 and 2000 cattle population in each of the DCL priority farming systems. The size of the cattle population generally follows the size of human population, the number of poor and the area of crops (Table 1, Table 4 and Table 7, respectively). Two farming systems stand out for their high population of cattle – rainfed mixed and rice-wheat, both in South Asia. However, high population and crop area in East Asia do not translate into the very high cattle populations seen in South Asia. For example, the three East Asia priority farming systems – lowland rice, upland intensive mixed and temperate mixed – have cattle populations in the middle of the range of the priority systems, although pig populations are larger. Other priority systems in the middle of the range include cereal-root crop mixed, agro-pastoral millet sorghum, pastoral, extensive cereal-livestock and maize mixed. While the small-scale cereal livestock system in Eastern Europe and Central Asia and the dry rainfed system in South Asia have relatively low cattle populations among the priority systems, livestock is clearly important in these systems.\n\nThe farming systems where dryland cereals and grain legumes are concentrated are particularly prone to drought and high temperatures (Table 5). Farming systems in areas with relatively low (and variable) annual precipitation are more susceptible to failed growing seasons, as shown in Figure 2. These dryland systems, especially those with less than 1000 mm of annual precipitation, tend to have a higher probability of drought or a failed season, when precipitation does not meet crop requirements. Areas that have high probabilities of being affected by drought as shown by the potential drought impact index (PDII) include the rainfed mixed system in South Asia and the agro-pastoral millet sorghum and pastoral systems in sub-Saharan Africa (Table 5). The rice-wheat system in South Asia also has a high PDII, where drought may particularly affect pearl millet and chickpea. Other systems that are particularly prone to drought include cereal-root crop mixed and maize mixed in sub-Saharan Africa and dry rainfed in South Asia.\n\nPriority dryland cereal and legume farming systems are labeled.\n\nThe DCL crops are also expected to be constrained by the rising temperatures that come with climate change. There is a general tendency of the drier farming systems having higher expected temperature changes between now and 2050 (Figure 3). Average temperature changes are expected to be between 2.4 and 3.4°C. The DCL priority farming systems in Eastern Europe and Central Asia could be particularly hard hit, with expected temperature rises of 3.3°C for extensive cereal livestock and 2.8°C for both large-scale cereal vegetable and small-scale cereal livestock. The estimated temperature change by 2050 in the temperate mixed system in East Asia is 2.9°C. For the rice-wheat system in South Asia and the agro-pastoral millet sorghum system in sub-Saharan Africa the expected change is 2.8°C, important expected changes because of their large area of DCL crop cultivation. The rice-wheat system has more than 11.2 million ha of DCL crops, while the agro-pastoral millet sorghum system has more than 18.6 million ha.\n\nPriority dryland cereal and legume farming systems are labeled.\n\nThe soils of DCL priority farming systems present a number of abiotic constraints to DCL crop production. Table 6 shows some of the principal constraints identified by the DCL commodity programs as they affect priority farming systems (DCL, 2015). The proportional area of farming systems with acid soil ranges from 9% in the small-scale cereal livestock system to 39% in the rainfed mixed system. The cereal-root crop mixed system is another one with a very large proportion – 37 percent – of its area exhibiting acid soils. Another system with a large area of acid soils is the maize mixed system in sub-Saharan Africa, with 27 percent of its area under this constraint. Lowland rice and upland intensive mixed in East Asia have nearly a quarter of their areas with acid soil constraints. These latter two systems also suffer from large areas with soils of low nutrient availability, with over one third of the area under this condition. Low nutrient availability is also an important constraint in the cereal-root crop mixed, maize beans and rainfed mixed systems, with proportional areas of 19, 14 and 10 percent of their total areas under this constraint, respectively. Salinity constraints are less problematic, with 13 of the 15 priority farming systems having less than 6% of their areas with this condition. The exceptions for soils with salinity constraints are the rice-wheat system in South Asia with 23 percent and the temperate mixed system in East Asia with 18 percent of their areas subject to salinity constraints. A group of farming systems has between 20 and 40 percent of their areas on soils with low moisture holding capacity – an important constraint in dryland systems due to the need for soils to store water for as long as possible. These systems include the agro-pastoral millet sorghum (38%), pastoral (30%) and cereal-root crop mixed (22%) farming systems in sub-Saharan Africa.\n\nThe key DCL farming system regions are home to about half of the global population, including a massive number of people living in poverty (Table 7). About 3.5 billion people live in these areas, 2.3 billion of them living in rural areas and 1.3 billion in towns and cities. The highest populations are in South Asia and East Asia. The lowland rice and upland intensive mixed systems in East Asia are two of the largest systems in terms of population, with roughly 851 and 501 million people in each respective system. Important South Asian farming systems include large numbers of urban and rural people – with over 400 million people in the rainfed mixed system and over 600 million people in the rice-wheat system. The remaining 14 DCL priority farming systems have a total of more than 960 million people.\n\nThe key DCL farming systems are home to about one third of the global population, including an enormous number of people living in poverty.\n\nThe DCL priority farming systems are home to a large proportion of the world’s poor (Table 7). According to year 2005 childhood stunting and $1 and $2 a day poverty indicators, about 60% of the world’s poor live within these 18 systems (Table 7; FAO, 2007; Stanley Wood, personal communication). This large proportion is due to the importance of these systems in high-population countries like China and India, as well as farming systems spanning West and East Africa. Of the 63 global farming systems, the DCL priority systems include eight of the top 10 systems in terms of numbers of poor people. These eight DCL systems are rice-wheat and rainfed mixed in South Asia, lowland rice, upland intensive mixed and temperate mixed in East Asia and cereal-root crop mixed, maize mixed and agro-pastoral millet sorghum in sub-Saharan Africa.\n\nUsing the population of stunted children as a nutrition and poverty indicator, more than 60 percent of the 2005 global population of stunted children live within the DCL priority farming systems (Table 7; de Onis et al., 2012; FAO, 2007). Much of this poverty is concentrated in South Asia, East Asia and sub-Saharan Africa, regions with historically high rates of malnutrition. According to the stunting indicator, two farming systems stand out, both in South Asia. The rice-wheat and rainfed mixed systems have 28 and 24 million stunted children, respectively, with malnutrition levels exemplifying the high population density and well-known nutrition problems of these regions. In the lowland rice and upland intensive mixed systems of East Asia, the number of stunted children is about half of the South Asian systems mentioned previously, with 13 and 15 million stunted children respectively. In sub-Saharan Africa, the maize mixed and agro-pastoral millet sorghum systems have about six million stunted children each, half again as much as the East Asia systems mentioned above. Another seven farming systems across five world regions have between one and four million stunted children. The remaining five farming systems regions – two in the Middle East and North Africa regions and three in the Eastern Europe and Central Asia regions – have less than one million stunted children, mostly reflecting the lower overall populations of these regions.\n\nThe DCL crops present a number of opportunities for bringing multiple technology options among different crops to the same geographic area (Figure 4). The map shows several core areas where three to five or more DCL crops are grown together. These core areas include (1) a large area spanning the Sahel region of West Africa, (2) a discontinuous cluster of areas in East Africa, (3) a large part of South Asia extending from India north to Pakistan and then east to Bangladesh and Myanmar, and (4) a large swath of area in the Middle East extending from Iran to Turkey. But there are also concentrations of multiple crops in Mexico and Central America, China and other regions. Figure 5 shows some of the crop combinations with the largest area. In the Sahel region, a huge area where groundnut, pearl millet and sorghum are grown together is found. The eastern part of this region contains systems that include these crops plus common bean, while the Western part of the region includes the same crops and much more cowpea cultivation. In the Middle East, the combination of barley, chickpea, lentil and faba bean, make up large cultivated areas within the region.\n\nThe map shows several core areas where 3 to 5 or more DCL crops are grown together.\n\n\nDiscussion and Conclusions\n\nThis study identified 18 farming systems globally that are important for dryland cereals and grain legumes agri-food systems. The most important of these systems, in terms of area and population, are found in South Asia and sub-Saharan Africa. The results discussed above suggests that these two regions deserve primary focus based on their relatively large cultivated area of DCL crops, large populations and high poverty. The farming systems in Latin America, the Middle East and North Africa, Central Asia and East Asia are also important. Research in any one region can have managed spillover effects in the others because the crops and biophysical conditions are similar across these regions. Interestingly, many DCL cereals and legumes show wide adaptability and have persisted in areas of moderate rainfall, such as the maize mixed farming system, and even in irrigated farming systems such as lowland rice.\n\nA focus on the 18 farming systems identified in this study in no way excludes any areas of the globe as areas where DCL should conduct and target research and development. It simply narrows down the DCL focus area to areas with substantial production of DCL commodities, with drylands and with substantial poverty and development problems.\n\nThis result can be compared against two existing maps, both of which can be viewed on the website of the DCL Atlas. One map shows dryland ecologies and another shows the countries prioritized by the DCL research program (DCL, 2015). The dryland ecologies map is solely based on dryness, as indicated by temperature, precipitation and evapotranspiration. Effectively, the map includes large areas where there are very few people and almost no cultivated land. The 18 farming systems identified in this research fall within the dryland ecologies map. Two partial exceptions to this pattern are the maize-beans system in Mesoamerica and the rainfed mixed system in India, where the boundaries of the farming system extend beyond the dryland ecology boundaries.\n\nDCL’s target countries map – based on national-level data – was developed using a combination of factors, namely, target crop area, agricultural population, population under poverty, prevalence of child malnutrition, and to the extent that data was available, land degradation based on the satellite-derived Normalized Difference Vegetation Index (NDVI; Pettorelli et al., 2005). The emphasis was on countries in dryland ecologies. In an effort to prioritize the large number of countries (51+), the focus was defined to be on sub-Saharan Africa and South Asia, where the area under the combined DCL was the highest among an assembled list of Low-Income Food-Deficit Countries (Pingali & Stringer, 2003). The target countries map also agrees well with the map of 18 farming systems. However, one drawback of the target countries approach is that it cannot distinguish between data representing the crop distribution and agroecology of DCL crops on the one hand, and country level data that was used for priority setting on the other. The results of this study overcome that obstacle by combining farming systems and countries, and by taking a more detailed spatial approach at subnational pixel level, as opposed to country level.\n\nThe results of this study can also be compared to a previous study that used the same approach, but with 23 crops, including the major staples rice, wheat, cassava and maize, among others (Hyman et al., 2008). That study focused on developing-country agriculture and prioritized 15 farming systems, with an emphasis on regions with large cultivated crop areas and large numbers of people. Seven of this study’s farming systems were not included in the previous study – pastoral, dry rainfed (SA), highland mixed (MENA), dryland mixed, large-scale cereal vegetable, small-scale cereal livestock and extensive cereal-livestock. These seven systems are mostly focused on DCL crops, have generally lower populations and cultivated areas, have a greater tendency towards mixed livestock and cereal production and are found in areas with lower rainfall. Five of the 15 systems in the previous study do not appear in this study on DCL. Two of these systems are lowland and very wet – rice in South Asia and root crop in sub-Saharan Africa. The other three systems in the previous study but not found in this one are highland systems in South Asia, sub-Saharan Africa and East Asia. The eight systems found in both studies show the importance of DCL crops to the global agricultural research and development effort. Six of the most important farming systems globally from the previous study (Hyman et al., 2008) are also systems important for DCL. They are rice-wheat and rainfed mixed in South Asia, cereal-root crop and maize mixed in sub-Saharan Africa and upland intensive mixed and lowland rice in East Asia. While the latter two systems have relatively small proportions of DCL crops, the absolute areas and benefiting populations are large in densely populated Southeast and East Asia.\n\nDryland cereal and legume crop distribution data show that South Asia and sub-Saharan Africa are the most important regions for crop improvement and adapted crop management practices. However, the proportional area of many DCL crops is often relatively low in regions where rice, wheat and maize are important staples. Nevertheless, the DCL crops are important in these regions for several reasons. Grain legumes in particular may be important as a rotation crop to support soil nitrogen fixation. Because livestock are important in many of the 18 farming system regions prioritized in this research, taking advantage of crop-livestock system synergies is an opportunity that should be explored, especially in relation to fodder. Also, the benefits of pasture and long term crop rotations in relation to soil improvement and reduction of plant disease can be considerable. The substantial ranges between yields in different regions of the world suggest considerable scope for closing yield gaps. These differences suggest substantial opportunities for increasing sub-Saharan Africa and South Asian yields.\n\nAbiotic constraints are significant obstacles to improving DCL production. Previous research showed that farmers in these DCL priority farming systems face potential drought conditions that have a much higher risk compared to most other farming systems (Hyman et al., 2008). Rising temperatures in DCL farming systems will place a growing demand on farmers to cultivate heat tolerant crops, and to develop practices to protect these crops. Farming systems on the edge of the tropics or in the subtropics, as one moves away from the equator, are more likely to face rising temperatures with climate change. The combined effects of drought and heat in these farming systems pose a significant challenge. The areas in the 18 priority farming systems show considerable soil limitations. One of the most important is infertility, as indicated by soil acidity and low nutrient reserves – for which legume crops are valuable. Other important soil limitations are related to water. Long dry seasons limit the water availability in the soil, which is compounded in coarse-textured soils with low water holding capacity, for which modern crop management practices are applicable. The dryness of these systems also make them susceptible to salinity, another important soil constraint in the DCL priority systems.\n\nSocioeconomic conditions in the DCL priority systems identified in this study indicate high levels of population and high poverty. There are both large rural and urban populations, suggesting potential positive supply and demand dynamics, especially so in sub-Saharan Africa and South Asia. These conditions suggest opportunities for developing market oriented production. Clearly much of the DCL crop production will continue to be derived from semi-subsistence agriculture. The high levels of malnutrition as indicated by childhood stunting, especially in South Asia and sub-Saharan Africa, can be addressed in part by nutritious DCL crops, which are often important sources of protein and micronutrients. Biofortification of DCL crops could be an important consideration in these areas. Clearly, the high rural and urban population found in, and depending on, DCL farming system regions suggest the importance of these systems for research and development aimed at improving agriculture and livelihoods.\n\nThe areas where DCL crops occur together present opportunities for improving the efficiency of research and development because fixed costs of research activities can be shared by different crop commodity programs. Testing the performance of crop varieties is typically carried out by national agricultural research institutes in collaboration with CGIAR centers. An integrated program to develop joint research could take advantage of different CGIAR centers or commodity programs carrying out research activities at common experiment stations of a national agricultural research institute. Two regions stand out where DCL crops occur together (Figure 5). An initiative to work on multiple crops in the same sites may be attractive for (1) millet, sorghum, groundnut and cowpea crops in the Sahel region, and for (2) barley, chickpea, lentil and faba bean in the Middle East.\n\nThis study points out several areas for further agricultural research to improve productivity of dryland cereals and legumes. First, an effort can be made to update this analysis using more recent data with higher temporal resolution. Using recent data is particularly important for crop distribution and socioeconomic data. This type of analysis will surely benefit from higher spatial resolution of geospatial data in the future, a trend increasingly common with improving capacities to collect, store and process geographic information. Perhaps the most substantial gap in this study has been the lack of information on biotic constraints to crop production. Pests and diseases are often the most important threats facing farmers. But to date there are few consistent and standardized geographic assessments of the major pests and disease threats to crops, notwithstanding the progress made on wheat rust prevalence (Singh et al., 2008). Overcoming this obstacle would require a systematic effort to collect information on the occurrence of biotic constraints. A recent paper showed the potential of improving our knowledge of the geographic dimensions of agricultural biodiversity (Castañeda et al., 2016). Interestingly, that research showed that the dryland systems area of the Middle East and North Africa is a priority for collecting wild relatives of food crops. Our research suggested the importance of temperature and precipitation under climate change for the future of DCL crops. Research is needed on understanding the sensitivity of each crop to increases in temperature and to the duration of drought conditions. Research is also needed on understanding genotype by environment interactions for the DCL crops. Other staple crops such as maize, wheat and rice have a better track record in these types of studies, suggesting a higher potential return on investment for this type of research on DCL crops in the future. In relation to comprehensive analyses which position DCL crops in the full farming systems in the two priority regions for DCL crops, a new edited volume will be valuable, a forthcoming book titled Farming Systems and Food Security in Africa: Priorities for Science and Policy Under Global Change, edited by John Dixon, Dennis Garrity, Jean-Marc Boffa, Tim Williams, Tilahun Amede with Christopher Auricht, Rosemary Lott and George Mburathi. A similar comprehensive analysis is recommended for South Asia.\n\n\nDeclarations\n\nThis research initiative developed two general types of data – digital spatial data of the world related to DCL crops and tabular data that summarizes the geographic information by farming system and country. All the spatial data used in the analysis can be accessed from the DCL Atlas at http://www.eatlasdcl.cgiar.org/. This online atlas includes data on the distribution of the 12 DCL crops, maps of predicted suitability of each crop, maps of abiotic constraints to crop production, maps of the biodiversity of relatives of each crop species, maps of socioeconomic conditions important for understanding the environment where these crops are grown and reference maps for putting all this information in the context.\n\nTabular data summarizing geographic information by farming system and country can be found in two dataset resources on the Dataverse repository:\n\nHarvard Dataverse: Replication Data for: Priority regions for research on dryland cereals and legumes, 10.7910/DVN/EDMXSK (Barona et al., 2016a).\n\nHarvard Dataverse: Characterization data on crops, production systems, abiotic constraints, population and poverty for farming system regions of the world, 10.7910/DVN/PLJ4SC (Barona et al., 2016b).",
"appendix": "Author contributions\n\n\n\nGH, EB and SB conceived the study. GH, EB, SB and SS designed the research. EB, CB, SC, HE, EG and OR prepared the datasets for processing and analysis. GH and JD wrote the manuscript. All authors interpreted and discussed the results and commented on the manuscript.\n\n\nCompeting interests\n\n\n\nThis research was commissioned by the CGIAR, a consortium of international research centers where all but one of the authors is employed. Future funding of research programs on dryland cereals and legumes could be influenced by the results and information found in this paper, which could impact on funds received by individual CGIAR centers. The authors declare no direct competing interest. All organizations for which the authors are employed are nonprofit organizations dedicated to developing global public goods.\n\n\nGrant information\n\nThis research was commissioned by the CGIAR Research Programs on Grain Legumes (http://grainlegumes.cgiar.org/) and on Dryland Cereals (http://drylandcereals.cgiar.org/). The CGIAR Fund (http://www.cgiar.org/who-we-are/cgiar-fund/) provided financial support for the study to the authors from four Consultative Group on International Agricultural Research (CGIAR) centers – CIAT, ICARDA, ICRISAT and IITA.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Liang You of the International Food Policy Research Institute (IFPRI) and Nora Castañeda-Álvarez of CIAT and the Crop Wild Relatives project for data on crop distributions and crop biodiversity, respectively, and for advice on the use of this data. We thank the CGIAR Research Program on Dryland Cereals and Legumes Agri-Food Systems (DCL), the International Center for Agricultural Research in the Dry Areas (ICARDA) and the International Center for Tropical Agriculture (CIAT) for providing in-kind support to this research.\n\n\nReferences\n\nBarona E, Guevara E, Hyman G: Replication Data for: Priority regions for research on dryland cereals and legumes. [online] Harvard Dataverse, V1 [UNF: 6:loJm7wy0VRiSBz6QuGnfbA==]. 2016a. Publisher Full Text\n\nBarona E, Guevara E, Hyman G: Characterization data on crops, production systems, abiotic constraints, population and poverty for farming system regions of the world. [online] Harvard Dataverse, V1. 2016b. Publisher Full Text\n\nCastañeda-Álvarez NP, Khoury CK, Achicanoy HA, et al.: Global conservation priorities for crop wild relatives. Nat Plants. 2016; 2: 16022. Publisher Full Text\n\nCIESIN (Center for International Earth Science Information Network), FAO (Food and Agriculture Programme), CIAT (Centro Internacional de Agricultura Tropical): Gridded Population of the World, Version 3 (GPWv3): Population Count Grid. [online] Columbia University, United Nations. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). 2005. Publisher Full Text\n\nCIESIN (Center for International Earth Science Information Network): Poverty Mapping Project: Global Subnational Prevalence of Child Malnutrition. Columbia University, Palisades NY, NASA Socioeconomic Data and Application Center (SEDAC). 2005. Reference Source\n\nCIESIN (Center for International Earth Science Information Network): Gridded Population of the World, Version 4 (GPWv4): Preliminary Release 2 (2010). [online] Columbia University Palisades, NY, 2014. Reference Source\n\nDCL (Dryland Cereals and Legumes): Dryland Cereals and Legumes Agri-food. Systems: Pre-proposal. 2015. Reference Source\n\nDESA (Department of Economic and Social Affairs): Rethinking Poverty: Report on the World Social Situation 2010. New York. 2009. Reference Source\n\nde Onis M, Blössner M, Borghi E: Prevalence and trends of stunting among pre-school children, 1990–2020. Public Health Nutr. 2012; 15(1): 142–148. PubMed Abstract | Publisher Full Text\n\nDixon J, Gulliver A, Gibbon D: Farming Systems and Poverty: Improving Farmers’ Livelihoods in a Changing World. FAO and World Bank, Rome and Washington DC, USA, 2001. Reference Source\n\nFischer G: Length of growing period data. Data not available for distribution. 2009. Reference Source\n\nFAO (Food and Agriculture Organization): Prevalence of stunting among children under five, by lowest available subnational Administrative unit, varying years (FGGD-Digital Atlas). Rome, 2007. Reference Source\n\nHartman GL, West ED, Herman TK: Crops that feed the World 2. Soybean—worldwide production, use, and constraints caused by pathogens and pests. Food Sec. 2011; 3(1): 5–17. Publisher Full Text\n\nHijmans RJ, Cameron SE, Parra JL, et al.: Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005; 25(15): 1965–1978. Publisher Full Text\n\nHyman G, Fujisaka S, Jones P, et al.: Strategic approaches to targeting technology generation: Assessing the coincidence of poverty and drought-prone crop production. Agr Syst. 2008; 98(1): 50–61. Publisher Full Text\n\nHyman G, Hodson D, Jones P: Spatial analysis to support geographic targeting of genotypes to environments. Front Physiol. 2013; 4: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHyman G, Fujisaka S, Jones P, et al.: Replication data for: Strategic approaches to targeting technology generation: Assessing the coincidence of poverty and drought-prone crop production. [online] Harvard Dataverse, V1. 2015. Publisher Full Text\n\nJones PG, Thornton PK: MarkSim: Software to generate daily weather data for Latin America and Africa. Agron J. 2000; 92(3): 445–453. Publisher Full Text\n\nJones PG, Thornton PK, Diaz W, et al.: MarkSim, Version 1. A computer tool that generates simulated weather data for crop modelling and risk assessment. Centro Internacional de Agricultura tropical (CIAT) CD-ROM series. Cali, Colombia: CIAT. CD-ROM + Guide. 2002; 87. Reference Source\n\nNewton AC, Flavell AJ, George TS, et al.: Crops that feed the world 4. Barley: a resilient crop? Strengths and weaknesses in the context of food security. Food Secur. 2011; 3(2): 141–178. Publisher Full Text\n\nPettorelli N, Vik JO, Mysterud A, et al.: Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends Ecol Evol. 2005; 20(9): 503–510. PubMed Abstract | Publisher Full Text\n\nPingali P, Stringer R: Food Security and Agriculture in the Low Income, Food-Deficit countries: 10 years after the Uruguay Round. 2003; (No. 03-18). Reference Source\n\nRamirez J, Jarvis A: High resolution statistically downscaled future climate surfaces. International Center for Tropical Agriculture (CIAT); CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS), Cali-Colombia. 2008. Reference Source\n\nRobinson TP, Wint GR, Conchedda G, et al.: Mapping the global distribution of livestock. PLoS One. 2014; 9(5): e96084. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanchez PA, Palm CA, Buol SW: Fertility Capability Soil Classification: a tool to help assess soil quality in the tropics. Geoderma. 2003; 114(3–4): 157–185. Publisher Full Text\n\nSingh RP, Hodson DP, Huerta-Espino J, et al.: Will stem rust destroy the world's wheat crop? Adv Agron. 2008; 98: 271–309. Publisher Full Text\n\nSmith RM: Comparing traditional methods for selecting class intervals on choropleth maps. Prof Geogr. 1986; 38(1): 62–67. Publisher Full Text\n\nYou L, Wood S: An entropy approach to spatial disaggregation of agricultural production. Agric Syst. 2006; 90(1–3): 329–347. Publisher Full Text\n\nYou L, Wood S, Wood-Sichra U, et al.: Generating global crop distribution maps: From census to grid. Agric Syst. 2014a; 127: 53–60. Publisher Full Text\n\nYou L, Wood-Sichra U, Fritz S, et al.: Spatial Production Allocation Model (SPAM) 2005 v2.0. 2014b. Reference Source\n\nZomer RJ, Trabucco A, Bossio DA, et al.: Climate change mitigation: A spatial analysis of global land suitability for clean development mechanism afforestation and reforestation. Agr Ecosyst Environ. 2008; 126(1–2): 67–80. Publisher Full Text"
}
|
[
{
"id": "13828",
"date": "25 May 2016",
"name": "Jeffrey D. Ehlers",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPositives: Using a basic farming system characterization schema developed some 15 years prior, this paper selects 18 of these where key dryland cereals and legume crops (DCL crops) are important, and refines their boundaries using new data and GIS resources. This work then provides a needed update to the earlier 2001 work. Each of these cropping system units is then characterized with DCL crop information (area grown, productivity, proportion of area, present and future drought severity, as well as socio-economic data such as population and poverty, as well as with cattle population data, to create an information rich view of these systems. This work therefore helps in discovering appropriate and high leverage investment opportunities that support economic development. Another value of the article is that it serves to publicize an open access dataset and information resources that can be used in future analysis of DCL systems.\nA valuable contribution of this work is the integration of datasets/information across the DCL crop, agro-ecological and socio-economic dimensions, especially at sub-national level. At the same time there is little doubt the data sets used in the analysis were of highly variable quality. It seems like it would be good to face this head-on and include more discussion around data quality issues and how much confidence the reader should put in the analysis given that some of the data likely to be of low to moderate quality. How are conclusions affected when integrating datasets on one topic that are of high quality with another for another parameter that are of poor quality?\nThe authors usefully point out some important research areas that need attention in the DCL crops to make the current maps more actionable. These include things like relative response to higher temperature, genotype by environment interactions but also that ‘one of the most important shortcomings in this type of analysis has been the lack of data with which to produce a biotic constraints map layer as these types of constraints can be key drivers of crop distribution.",
"responses": []
},
{
"id": "14209",
"date": "06 Jun 2016",
"name": "Liangzhi You",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nEfficient allocation of limited research resources is a central policy issue for international development agencies, developing-country governments, and their agricultural research institutes. Priority setting in agricultural research systems is a principal means of ensuring the effectiveness of investment in agricultural research. The new CGIAR has emphasize more on research impact on development. As part of the CGIAR research program on dryland cereal and legumes, this paper developed an open access dataset and information resource that provides the basis for future analysis of the geographic dimensions of dryland cereals and legumes.\n\nThe foremost important contribution of this paper is to integrate multiple global spatial databases on cropping system, livestock and biophysical conditions for setting the priority. These newly available datasets allows the priority setting at crop-type level in a truly spatially explicit manner. The use of geographic information system (GIS) and spatial analysis captured the effects of agroecological variation and huge spatial heterogeneity of farming system, the response of agricultural technologies, the technological spillovers.\n\nThe next strength of the paper is its open access. The paper itself is open access. The datasets used in the paper is also openly available. Interested users could download these datasets and repeat/improved the analysis undertook in the paper.\n\nThe weakness is the methodology section. While the paper spent quite some effort/sections on data sources, summary of results, the methodology is quite short and not clear. Admittedly the readers are referred to the well-known farming system framework by Dixon et al., and the authors’ previous work. The methodology should be updated and improved with the new datasets and tools. A brief summary is warranted. Technical details on how to deal with spatial fragmentation (e.g due to many different data layers) would be helpful for those who would like to learn and repeat the analysis of this paper.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-885
|
https://f1000research.com/articles/5-1595/v1
|
06 Jul 16
|
{
"type": "Review",
"title": "Adolescent Klinefelter syndrome: is there an advantage to testis tissue harvesting or not?",
"authors": [
"Robert Oates"
],
"abstract": "It is currently unclear whether an adolescent with 47,XXY Klinefelter syndrome will be better off having testicular sperm extraction (TESE) performed in an effort to ‘preserve fertility’ for the future or, alternatively, should be advised to simply wait until adulthood when he and his partner are ready to begin a family. This report will provide data suggesting that there is no obvious ‘preservation’ benefit and that recommending TESE to the 47,XXY boy and his parents may not be as helpful as it might appear and may be overly aggressive.",
"keywords": [
"testicular sperm extraction",
"fertility",
"non-obstructive azoospermic",
"spermatogenesis"
],
"content": "Introduction\n\nKlinefelter syndrome (KS) (47,XXY) is the most common chromosomal disorder in men, affecting one in 600 newborn males, and is etiologic in up to 11% of non-obstructive azoospermic (NOA) males1. When testicular sperm extraction (TESE) became a standard technique for the discovery and retrieval of sperm in the NOA patient, it was subsequently successfully applied to KS adults, who had previously been considered sterile. Published reports from many different centers describe rates of sperm retrieval in these 47,XXY adults to be approximately 50–70%, along with excellent pregnancy rates and healthy 46,XY or 46,XX offspring, as recently reviewed by Majzoub et al.2.\n\nFor a variety of reasons, KS may occasionally be diagnosed prenatally, in infants, in young boys, or in adolescent males3–7. Based upon the knowledge that not all adult KS males will have spermatozoa harvested upon TESE, several clinical investigators have pursued a course of action to determine if TESE should be offered and applied to the 47,XXY adolescent in order to “preserve fertility potential”8,9. This submission will address this vexing question from many different angles and try to answer whether this is an overly aggressive approach with more harm than benefit or if this is an appropriate strategy with more benefit than harm.\n\nThe underlying tenets for recommending TESE in adolescent KS males is that fertility potential and the presence of testicular spermatozoa are maximal at puberty or just after, and that both are precipitously and irrevocably lost over the next several years. Is that actually true or has fiction become belief that has become fact that has become dogma? Additionally, some feel that testosterone levels plummet after puberty and require replacement, and TESE should be carried out prior to that. However, in 1985, Saltenblatt concluded that testosterone levels settle at low-normal values for most with KS after puberty and remain at those levels into adulthood10. More recently, Aksglaede et al. reviewed 166 non-mosaic KS boys and demonstrated that there is no rapid, inexorable decline in their serum testosterone values (which were also low-normal in most subjects but high-normal in some) spanning early adolescence to adulthood11.\n\n\nWhy is there spermatogonial cell apoptosis at puberty? Is it helpful to preserve these cells?\n\nAt conception, the nascent embryo is 47,XXY and all cells, including those destined to become gonocytes and eventually spermatogonial stem cells, will have this same chromosomal constitution. As the spermatogonial stem cell precursors and stem cells themselves migrate to the gonadal ridge from the yolk sac, they expand mitotically in number and begin the slow process of differentiation12–14. Even after invading and populating the emergent seminiferous tubules, numerical increase continues until birth. With the onset of the mini-puberty (neonatal surge of gonadotropins) during the first few months of life, proliferation and differentiation (of some) to type A dark (Ad) spermatogonia commences and, when this temporally limited hypothalamic-pituitary stimulation stops, the spermatogonia become quiescent until puberty, although there may be a gradual diminution in the absolute numbers of spermatogonia in the first year of life15,16. The vast majority of these resting cells would be 47,XXY but, occasionally, it is thought the supernumerary X chromosome is lost during an earlier mitotic replication and the resultant spermatogonial stem cell or type Ad spermatogonia is, therefore, normally diploid (46,XY)17. These are thought by many to be the cells that eventually, upon initiation of puberty, will be capable of completing the full process of spermatogenic differentiation (mitosis, meiosis, and spermiogenesis), their progeny being fully functional haploid spermatozoa. However, the more numerous, by orders of magnitude, 47,XXY spermatogonia suffer a meiotic block, arrest, and become apoptotic18. Whether it is simply the trisomic state or the overexpression of X-linked, testis-expressed genes that lead to this demise is unclear, although the latter hypothesis has more evidence behind it19,20. The legacy of this self-destruction is wide swaths of the testicular parenchyma with seminiferous tubules that are empty ghosts or unrecognizable and fibrotic. By happenstance, every once in a while, a seminiferous tubule in which a 46,XY spermatogonium found itself at home survives, incubates, and cultivates the normal machinery responsible for complete spermatogenesis21. So, if 47,XXY spermatogonia have no ability to birth whole spermatozoa, is there a reason to harvest and cryopreserve them before they become apoptotic early in puberty? The answer would appear to be “no”, as concluded by Oates22. Indeed, with regard to fertility preservation in even younger males with KS, as Gies et al. cautiously posit, “given these controversies, banking testicular tissue from prepubertal KS boys should be performed only in a research framework”23, even though parents of KS boys would be in favor of it24.\n\n\nIs there a better chance of finding spermatozoa in adolescence than in adulthood?\n\nGiven the above discussion that spermatozoa arise from random 46,XY spermatogonia scattered about a sea of fibrotic tubules, the next obvious question in the search to answer whether it is advantageous to perform TESE in an adolescent KS male as opposed to waiting until adulthood would naturally be whether the chance of finding sperm upon TESE is greater in the adolescent than it is in the adult. If so, that would be a rationale for TESE in these younger males, but if not there would seem to be no benefit in doing so. That is, is there a great likelihood that these competent and capable 46,XY populated seminiferous tubules will also disappear during puberty or in the next several years, perhaps collateral damage of the near total annihilation of the neighboring 47,XXY spermatogonia and their home tubules and, thusly, they and any sperm they produce should be harvested and saved for the future as soon as possible?\n\nSperm seen within the ejaculate are probably only a small percentage of the total number produced in the testis when viewing the entirety of the testis parenchyma as a single manufacturing unit. When that output falls below a certain minimum number, but not to zero, no sperm can be found downstream in the seminal fluid, but some of the relatively few that have been created may still be identified when combing through and dissecting that testis tissue factory itself. Shortly following the introduction of intracytoplasmic sperm injection (ICSI), it was realized that testicular sperm was capable of fertilization, embryo development, and pregnancy. TESE quickly became the standard therapy offered to men with NOA of all types in an effort to find and use individual spermatozoa to achieve male genetic parenthood25–31. Retrieved spermatozoa could even be intentionally frozen prior to use in conjunction with ICSI, as first demonstrated by Oates et al., and this is now routine clinical practice32. In their recent systematic review and meta-analysis, Bernie et al. concluded that the sperm retrieval rate (SRR) was higher with micro-TESE (involving the use of an operating microscope) than with conventional TESE33 and approximates 50–60%, depending upon the center and the surgeon. Tournaye et al. initially pioneered its application in adult KS men, and the first pregnancy was documented by Palermo et al.34,35. Numerous subsequent publications confirmed this initial proof of concept and added to the growing cumulative statistic of approximately 50–60% likelihood of sperm retrieval in this subpopulation of NOA men2.\n\nDamani et al. first reported in 2001 the use of TESE in an adolescent male in order to ‘preserve’ his fertility before he was started on spermatogenically suppressive testosterone replacement36. The authors carefully presented their data and the decisions that led the patient and his parents to choose this approach in a cautionary tone, advocating only that TESE in the adolescent KS male needed more study and consideration before becoming viewed as routine, always beneficial, and standard. Building upon that idea, a handful of subsequent reports presented data on testis biopsy or TESE in very young/adolescent KS boys with varying results. In 2004, Wikstrom et al. reported that they were not able to find any spermatozoa in the histological preparations of single-site biopsies on 14 boys aged 10–14 years37. They noted a minimum number of type A pale (Ap) and Ad spermatogonia in seven of the 14, all within the age range of 10–12.5 years, but no spermatogonia of any kind in the remaining seven who had an age range of 11.7–14 years. In 2012, Gies and colleagues, who also performed a single-site biopsy but of “large volume”, reported that they too could not document full spermatogenesis on histological analysis in seven KS boys, aged 10.2–15.6 years38. Spermatogonia were seen in four of the seven in the upper age range of their group (13.3–15.9). Another seemingly disappointing report by Rivas et al. in 2013 stated that only one KS adolescent out of five operated upon had spermatozoa visualized, even though they carried out single-site bilateral TESE39. Mehta et al., also in 2013, using a microsurgical TESE approach reported positive news with sperm discovery in seven of 10 adolescents, ages ranging from 14–22 years: an SRR of 70%40. Plotten et al. in 2015 described an SRR of 52% in “young” KS patients, aged 15–239. Finally, Nahata et al. reported a 50% SRR in 10 adolescent KS boys aged 15–23 when performing microTESE41. Sperm was identified in five KS boys (aged 16, 16, 17, 19, and 23) but was not discovered in the remaining five KS boys (aged 15, 16, 16, 18, and 20), indicating no real pattern or prediction based upon age at the time of TESE.\n\nAs above, SRR rates of TESE in KS adults approximate 50–60%. But are there data to compare within the same program, adolescent to adult? Plotten et al. performed exactly that study showing prospectively, as detailed previously, an SRR of 52% in their “young” group and 62% (no statistical difference) in their “adult” group, aged over 23 years9. At Weill Cornell Medical College, Dabaja and Schlegel confined their reported results to 127 adult KS men and showed an SRR of 65%, which is similar to that of 70% (7/10) reported by Mehta et al. at the same institution40,42. As stated and suggested by Nahata et al. in their conclusion, “…there is no clear benefit to performing TESE/unilateral microTESE in adolescence…”41. In total, then, although perhaps intuitively satisfying, sensible, and gratifying, SRRs in the small numbers of adolescent KS boys subjected to TESE appear to be the same as those SRRs in adult KS men subjected to TESE. In general, the sperm retrieved from the adult is used immediately as the source of sperm for ICSI and not cryopreserved. Will cryopreserved spermatozoa or testis tissue be equally as useable and successful after years of storage as would necessarily be the case if harvested from a young boy not ready to conceive for years? Future reports detailing the success of using this sperm (obtained many years previously) will be necessary to answer this question. This will include those adolescents who have had microTESE prior to the institution of spermatogenically suppressive exogenous testosterone therapy.\n\n\nDoes sperm production decline in the years of adolescence and on into adulthood to negligible levels?\n\nThis is another way of asking the same question: should we be performing TESE in the adolescent to preserve fertility that will be irrevocably lost as the years pass and we do nothing? This notion also arises from very limited data suggesting that sperm production, as determined by results of TESE (namely SRR), declines after a certain age (35 or so) and as such it might be best to harvest tissue as soon as the diagnosis is made, whatever the age43. Although these conclusions have been made based on a limited data set from Okada et al.43 (25 patients greater than age 35 with a 23% SRR, far different from the Cornell group as detailed below), this is by no means certain. As a matter of fact, the data from the Cornell group44 show that the SRR in KS men was 71% in the 22–30 year age group, 86% in the 31–35 year age group, and 50% in the 36–50 year age group. The group age intervals are not of the same time spans, and the SRR is higher when you are 33 than when you are 26 (how is that possible biologically?) and is still 50% when greater than age 36. As they state, “In addition, ROC analysis of these data yielded an AUC of 0.58, indicating that there is no best age to predict SRR based on our data”. So it appears that even in the publications discussing whether or not TESE rates are less in the ‘older’ adult than in the ‘younger’ adult (aged 35), there is still a rate of success that is close to, and not much better than, that claimed for adolescents.\n\nFinally, when comparing overall adolescent rates to overall adult rates of TESE success, and although difficult to prove, there would be little reason to believe that sperm production would magically begin years after the onset of puberty, e.g. age 16 or 17. Therefore, since 50 out of every 100 KS boys have sperm recovered in their testis tissue and 50 out of every 100 KS adults have sperm recovered from their testis tissue, and if a decline in sperm production occurred during this time frame to zero values of, for argument’s sake, 10%, then 10% of maturing KS males (adolescent to adult ages) would have to begin to produce sperm to keep the two groups equivalent, as far as SRR rates go. Is it likely that 10 (or whatever absolute number is chosen) males have a precipitous and total decline in sperm production as they advance from adolescence into adulthood and, to keep the rates of what we actually find upon TESE balanced, 10 males begin to suddenly produce sperm as they advance from adolescence into adulthood? Or is it more likely that those who have seminiferous tubules populated by 46,XY spermatogonia capable of undergoing the full and complete sequence of spermatogenesis maintain those same tubules throughout the years of adolescence and on into adulthood? It may be the exact same individuals who have spermatozoa found in adolescence that are those found in adulthood. Where would the benefit be? Harm could certainly come to the adolescent and his parents when no sperm are retrieved—a potential heavy burden to carry in the difficult teenage years.\n\n\nConclusion\n\nThe basic biology of spermatogenesis and how it is altered in the KS testis is important to appreciate when trying to formulate appropriate clinical pathways vis-à-vis fertility and biological paternity. Although it seems obvious to some that KS boys benefit or even require testis tissue surgery and freezing of any sperm found as soon as they are identified in the adolescent years, the data do not support the temporal necessity to do so. It may be just as reasonable, if not more so, to perform that same surgery at a later date, in adulthood when the patient and his partner can make an informed choice as to whether an operation on him and an in vitro cycle for her is what they would like to do in an effort to build a family. Since the incidence of KS syndrome is 1:600, practitioners of all types are going to have many KS males as their patients, will be involved in this debate and controversy, and will need to be aware of the issues involved. This list includes, but is not limited to, obstetricians (amniocentesis results), pediatricians (discovered at the time of a learning disability evaluation, or to understand the reason for small testes), endocrinologists (detected during exploration for hypogonadism and failure to initiate puberty in the most severely affected), urologists, and reproductive endocrinologists (elucidated at the time of an infertility work-up). As it usually is, the final answer will probably lie somewhere in the middle and be determined on an individual patient-specific basis.",
"appendix": "Competing interests\n\n\n\nThe author declares that he has no competing interests.\n\n\nGrant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nReferences\n\nWikstrom AM, Dunkel L: Klinefelter syndrome. Best Pract Res Clin Endocrinol Metab. 2011; 25(2): 239–50. PubMed Abstract | Publisher Full Text\n\nMajzoub A, Arafa M, Al Said S, et al.: Outcome of testicular sperm extraction in nonmosaic Klinefelter syndrome patients: what is the best approach? Andrologia. 2016; 48(2): 171–6. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nHerlihy AS, Halliday JL, Cock ML, et al.: The prevalence and diagnosis rates of Klinefelter syndrome: an Australian comparison. Med J Aust. 2011; 194(1): 24–8. PubMed Abstract\n\nManning MA, Hoyme HE: Diagnosis and management of the adolescent boy with Klinefelter syndrome. Adolesc Med. 2002; 13(2): 367–74, viii. PubMed Abstract\n\nMehta A, Mielnik A, Schlegel PN, et al.: Novel methylation specific real-time PCR test for the diagnosis of Klinefelter syndrome. Asian J Androl. 2014; 16(5): 684–8. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nRadicioni AF, Ferlin A, Balercia G, et al.: Consensus statement on diagnosis and clinical management of Klinefelter syndrome. J Endocrinol Invest. 2010; 33(11): 839–50. PubMed Abstract | Publisher Full Text\n\nTincani BJ, Mascagni BR, Pinto RD, et al.: Klinefelter syndrome: an unusual diagnosis in pediatric patients. J Pediatr (Rio J). 2012; 88(4): 323–7. PubMed Abstract | Publisher Full Text\n\nMehta A, Paduch DA: Klinefelter syndrome: an argument for early aggressive hormonal and fertility management. Fertil Steril. 2012; 98(2): 274–83. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPlotton I, Giscard d'Estaing S, Cuzin B, et al.: Preliminary results of a prospective study of testicular sperm extraction in young versus adult patients with nonmosaic 47,XXY Klinefelter syndrome. J Clin Endocrinol Metab. 2015; 100(3): 961–7. PubMed Abstract | Publisher Full Text\n\nSalbenblatt JA, Bender BG, Puck MH, et al.: Pituitary-gonadal function in Klinefelter syndrome before and during puberty. Pediatr Res. 1985; 19(1): 82–6. PubMed Abstract | Publisher Full Text\n\nAksglaede L, Skakkebaek NE, Almstrup K, et al.: Clinical and biological parameters in 166 boys, adolescents and adults with nonmosaic Klinefelter syndrome: a Copenhagen experience. Acta Paediatr. 2011; 100(6): 793–806. PubMed Abstract | Publisher Full Text\n\nDavis SM, Rogol AD, Ross JL: Testis Development and Fertility Potential in Boys with Klinefelter Syndrome. Endocrinol Metab Clin North Am. 2015; 44(4): 843–65. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nHughes IA: Minireview: sex differentiation. Endocrinology. 2001; 142(8): 3281–7. PubMed Abstract | Publisher Full Text\n\nMagnon C, Lucas D, Frenette PS: Trafficking of stem cells. Methods Mol Biol. 2011; 750: 3–24. PubMed Abstract | Publisher Full Text\n\nMikamo K, Aguercif M, Hazeghi P, et al.: Chromatin-positive Klinefelter's syndrome. A quantitative analysis of spermatogonial deficiency at 3, 4, and 12 months of age. Fertil Steril. 1968; 19(5): 731–9. PubMed Abstract | Publisher Full Text\n\nMüller J, Skakkebaek NE, Ratcliffe SG: Quantified testicular histology in boys with sex chromosome abnormalities. Int J Androl. 1995; 18(2): 57–62. PubMed Abstract | Publisher Full Text\n\nVialard F, Bailly M, Bouazzi H, et al.: The high frequency of sperm aneuploidy in Klinefelter patients and in nonobstructive azoospermia is due to meiotic errors in euploid spermatocytes. J Androl. 2012; 33(6): 1352–9. PubMed Abstract | Publisher Full Text\n\nWikström AM, Dunkel L: Testicular function in Klinefelter syndrome. Horm Res. 2008; 69(6): 317–26. PubMed Abstract | Publisher Full Text\n\nJohannisson R, Gropp A, Winking H, et al.: Down's syndrome in the male. Reproductive pathology and meiotic studies. Hum Genet. 1983; 63(2): 132–8. PubMed Abstract | Publisher Full Text\n\nRoss MT, Grafham DV, Coffey AJ, et al.: The DNA sequence of the human X chromosome. Nature. 2005; 434(7031): 325–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSciurano RB, Luna Hisano CV, Rahn MI, et al.: Focal spermatogenesis originates in euploid germ cells in classical Klinefelter patients. Hum Reprod. 2009; 24(9): 2353–60. PubMed Abstract | Publisher Full Text\n\nOates RD: The natural history of endocrine function and spermatogenesis in Klinefelter syndrome: what the data show. Fertil Steril. 2012; 98(2): 266–73. PubMed Abstract | Publisher Full Text\n\nGies I, Oates R, De Schepper J, et al.: Testicular biopsy and cryopreservation for fertility preservation of prepubertal boys with Klinefelter syndrome: a pro/con debate. Fertil Steril. 2016; 105(2): 249–55. PubMed Abstract | Publisher Full Text\n\nGies I, Tournaye H, De Schepper J: Attitudes of parents of Klinefelter boys and pediatricians towards neonatal screening and fertility preservation techniques in Klinefelter syndrome. Eur J Pediatr. 2016; 175(3): 399–404. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCraft I, Bennett V, Nicholson N: Fertilising ability of testicular spermatozoa. Lancet. 1993; 342(8875): 864. PubMed Abstract | Publisher Full Text\n\nDevroey P, Liu J, Nagy Z, et al.: Pregnancies after testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermia. Hum Reprod. 1995; 10(6): 1457–60. PubMed Abstract | Publisher Full Text\n\nLewin A, Weiss DB, Friedler S, et al.: Delivery following intracytoplasmic injection of mature sperm cells recovered by testicular fine needle aspiration in a case of hypergonadotropic azoospermia due to maturation arrest. Hum Reprod. 1996; 11(4): 769–71. PubMed Abstract | Publisher Full Text\n\nMulhall JP, Burgess CM, Cunningham D, et al.: Presence of mature sperm in testicular parenchyma of men with nonobstructive azoospermia: prevalence and predictive factors. Urology. 1997; 49(1): 91–5; discussion 95–6. PubMed Abstract | Publisher Full Text\n\nMulhall JP, Reijo R, Alagappan R, et al.: Azoospermic men with deletion of the DAZ gene cluster are capable of completing spermatogenesis: fertilization, normal embryonic development and pregnancy occur when retrieved testicular spermatozoa are used for intracytoplasmic sperm injection. Hum Reprod. 1997; 12(3): 503–8. PubMed Abstract | Publisher Full Text\n\nSchoysman R, Vanderzwalmen P, Nijs M, et al.: Pregnancy after fertilisation with human testicular spermatozoa. Lancet. 1993; 342(8881): 1237. PubMed Abstract | Publisher Full Text\n\nSchoysman R, Vanderzwalmen P, Nijs M, et al.: Successful fertilization by testicular spermatozoa in an in-vitro fertilization programme. Hum Reprod. 1993; 8(8): 1339–40. PubMed Abstract\n\nOates RD, Mulhall J, Burgess C, et al.: Fertilization and pregnancy using intentionally cryopreserved testicular tissue as the sperm source for intracytoplasmic sperm injection in 10 men with non-obstructive azoospermia. Hum Reprod. 1997; 12(4): 734–9. PubMed Abstract | Publisher Full Text\n\nBernie AM, Mata DA, Ramasamy R, et al.: Comparison of microdissection testicular sperm extraction, conventional testicular sperm extraction, and testicular sperm aspiration for nonobstructive azoospermia: a systematic review and meta-analysis. Fertil Steril. 2015; 104(5): 1099–103.e1-3. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPalermo GD, Schlegel PN, Sills ES, et al.: Births after intracytoplasmic injection of sperm obtained by testicular extraction from men with nonmosaic Klinefelter's syndrome. N Engl J Med. 1998; 338(9): 588–90. PubMed Abstract | Publisher Full Text\n\nTournaye H, Staessen C, Liebaers I, et al.: Testicular sperm recovery in nine 47,XXY Klinefelter patients. Hum Reprod. 1996; 11(8): 1644–9. PubMed Abstract | Publisher Full Text\n\nDamani MN, Mittal R, Oates RD: Testicular tissue extraction in a young male with 47,XXY Klinefelter's syndrome: potential strategy for preservation of fertility. Fertil Steril. 2001; 76(5): 1054–6. PubMed Abstract | Publisher Full Text\n\nWikström AM, Raivio T, Hadziselimovic F, et al.: Klinefelter syndrome in adolescence: onset of puberty is associated with accelerated germ cell depletion. J Clin Endocrinol Metab. 2004; 89(5): 2263–70. PubMed Abstract | Publisher Full Text\n\nGies I, de Schepper J, van Saen D, et al.: Failure of a combined clinical- and hormonal-based strategy to detect early spermatogenesis and retrieve spermatogonial stem cells in 47,XXY boys by single testicular biopsy. Hum Reprod. 2012; 27(4): 998–1004. PubMed Abstract | Publisher Full Text\n\nRives N, Milazzo JP, Perdrix A, et al.: The feasibility of fertility preservation in adolescents with Klinefelter syndrome. Hum Reprod. 2013; 28(6): 1468–79. PubMed Abstract | Publisher Full Text\n\nMehta A, Bolyakov A, Roosma J, et al.: Successful testicular sperm retrieval in adolescents with Klinefelter syndrome treated with at least 1 year of topical testosterone and aromatase inhibitor. Fertil Steril. 2013; 100(4): 970–4. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nNahata L, Yu RN, Paltiel HJ, et al.: Sperm Retrieval in Adolescents and Young Adults with Klinefelter Syndrome: A Prospective, Pilot Study. J Pediatr. 2016; 170: 260–265.e2. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nDabaja AA, Schlegel PN: Microdissection testicular sperm extraction: an update. Asian J Androl. 2013; 15(1): 35–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOkada H, Goda K, Yamamoto Y, et al.: Age as a limiting factor for successful sperm retrieval in patients with nonmosaic Klinefelter's syndrome. Fertil Steril. 2005; 84(6): 1662–4. PubMed Abstract | Publisher Full Text\n\nRamasamy R, Ricci JA, Palermo GD, et al.: Successful fertility treatment for Klinefelter's syndrome. J Urol. 2009; 182(3): 1108–13. PubMed Abstract | Publisher Full Text | F1000 Recommendation"
}
|
[
{
"id": "14814",
"date": "06 Jul 2016",
"name": "Peter Schlegel",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14813",
"date": "06 Jul 2016",
"name": "Stanton Honig",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14812",
"date": "06 Jul 2016",
"name": "Richard N. Yu",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1595
|
https://f1000research.com/articles/5-1590/v1
|
06 Jul 16
|
{
"type": "Review",
"title": "Emerging concepts for PI3K/mTOR inhibition as a potential treatment for osteosarcoma",
"authors": [
"Michael W. Bishop",
"Katherine A. Janeway",
"Michael W. Bishop"
],
"abstract": "Patients with metastatic and recurrent osteosarcoma fare poorly, and new therapeutic strategies are needed to improve survival. Several recent complementary genomic and pathway analyses of both murine and human osteosarcoma have revealed common aberrations of the phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) pathway in osteosarcoma. Preclinical data demonstrate that inhibition of PI3K and mTOR with either a combination of single agents or dual inhibiting compounds can decrease cell proliferation and induce cell cycle arrest and apoptosis. With a lack of available clinical agents active in osteosarcoma, PI3K/mTOR inhibition represents a potential vulnerability in osteosarcoma that warrants clinical investigation.",
"keywords": [
"osteosarcoma",
"PI3K",
"mTOR",
"inhibition treatment"
],
"content": "Introduction\n\nOsteosarcoma is the most common malignancy of bone diagnosed in children and adolescents, with an age-adjusted incidence of 4.4 new cases per million each year1. Contemporary studies estimate that with the use of surgery and multimodal chemotherapy with high-dose methotrexate, cisplatin, and doxorubicin, 65 to 70% of patients are able to achieve long-term cure2. However, patients who present with metastatic disease fare poorly, with survival rates less than 30%3–5. Furthermore, for patients who relapse, survival is less than 20%, and cure is nearly impossible if surgical complete remission cannot be achieved6,7. Several avenues for augmenting treatment are currently under preclinical or clinical investigations, supported by recent biologic and genomic data. Among these, interest has been raised for the use of drugs targeting the phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) pathway as a potential vulnerability. Aberrations of this pathway have previously been described in osteosarcoma, such as PTEN deletion8 and PIK3CA mutations9, but were observed at relatively low frequency. However, contemporary biologic studies now reveal more frequent alterations of this pathway. The following brief review will highlight recent analyses supporting the role of PI3K/mTOR inhibition as an area ripe for further exploration in osteosarcoma.\n\n\nGenetic studies reveal potential role for PI3Km/TOR inhibition in osteosarcoma\n\nAs part of efforts to identify targets for novel therapeutic agents, institutional and collaborative group efforts to characterize the genomic landscape of osteosarcoma have been conducted with the hope of identifying targetable recurrent aberrations. Numerous genomic and epigenetic analyses have revealed the striking genomic complexity and heterogeneity among osteosarcoma samples but have also elucidated a few common themes including alterations of TP53 and/or RB1 in most samples, and distinct chromosomal regions of hypermutation (“kataegis”)10,11. However, a recent complementary genomic and pathway analysis identified PI3K/mTOR pathway aberrations in a subset of osteosarcoma samples. Heuristic analysis of whole genome, exome, and RNA sequencing data from 59 osteosarcoma tumors revealed alterations in the PI3K/mTOR pathway in 24% of samples that included aberrations of PTEN, TSC2, PIK3R1, PIK3CA, and several other genes. Using a comparative oncology approach, whole exome sequencing of a Tp53/Rb1 conditionally deleted osteosarcoma mouse model, somatic mutations in PTEN and PIK3R1 were observed in both murine and human tumors. Furthermore, a genome-wide shRNA screen of a primary murine osteosarcoma cell line identified 172 enriched genes including Pik3ca and Mtor, with inhibition of murine osteosarcoma cells11. Based on this information, two dual PI3K/mTOR inhibitors (GSK2126458, BEZ-235) and a PIK3CA-specific inhibitor (PIK75) were tested against human and murine-derived cell lines. All three drugs inhibited cell proliferation in all cell lines; PIK75 and GSK2126458 induced apoptosis as demonstrated by caspase 3/7 activation and poly(ADP-ribose) polymerase (PARP) cleavage.\n\nIn a separate systematic analysis, whole-genome siRNA screening of primary osteosarcoma cell cultures derived from a genetically engineered murine model revealed enrichment in pathways associated with protein translation and mTOR signaling. A small molecule/kinase inhibitor screen of murine-derived osteosarcoma cell lines revealed activity in compounds (PIK-75, GSK2126458, and BEZ-235) targeting PI3K and mTOR and/or DNA-PK. Activity of dual PI3K/mTOR inhibitors was subsequently observed in cell death assays of cultures from primary human xenograft-derived osteosarcoma (GSK2126458, PKI-587, BEZ-235, and BGT-226). Administration of the compounds GSK2126458 and PKI-587 inhibited phosphorylation of downstream targets in a dose-dependent manner, increased the number of cells in the G0-G1 phase, and induced apoptosis in both murine and human cell lines. Combinations of PI3K- or mTOR-specific inhibitors were also evaluated, and while individual activity was not observed, combination of the PIK3CA-specific inhibitor BYL719 and everolimus yielded a synergistic interaction12.\n\nFurther supporting these comprehensive analyses, the use of novel genetic screening technologies provides additional evidence for the importance of the PI3K/mTOR pathway in osteosarcoma. A Sleeping Beauty (SB) transposon-based forward genetic screen was performed in mice with and without somatic loss of Trp53 to identify common insertion sites associated with the development of osteosarcoma. Pten was one of the most commonly mutated genes in both Trp53-SB-mutated and non-Trp53-SB-mutated tumors. Nf2 and Nf1, both of which serve regulatory functions for downstream mTOR signaling, were also frequently mutated in the SB-mutated tumors. Pathway analysis identified enrichment for candidate genes in the PI3K/AKT/mTOR pathway, as well as overlap with the ErbB and ERK/MAPK pathways. Furthermore, conditional knockdown of both Trp53 and Pten in a mouse model accelerated the development of osteosarcoma, and knockout of PTEN in an immortalized osteoblast cell line with inhibited TP53 function led to significantly increased colony formation, suggesting that PTEN loss is cooperative with TP53 dysfunction to drive osteosarcomagenesis and proliferation13.\n\nSeveral clinical reports describing activity of agents targeting mTORC1/2 have included patients with osteosarcoma. A recent report from the French Sarcoma Group of off-label use of targeted therapies for osteosarcoma found that those who received rapamycin (with or without cyclophosphamide) compared to a group of tyrosine kinase inhibitors (sunitinib, sorafenib, and pazopanib) had a superior progression-free survival (PFS) (hazard ratio [HR] 2.7, 95% confidence interval [CI] 1.05–7.1), although the difference in median PFS was modest (3 months vs. 1.8 months)14. A phase II study of the mTOR inhibitor ridaforolimus included two osteosarcoma patients with a confirmed partial response and one patient with an unconfirmed partial response15. Fifty osteosarcoma patients were enrolled on a subsequent phase III study using ridaforolimus as a maintenance therapy but were included in a cohort of bone tumors and were not analyzed separately. In a subgroup analysis, the use of ridaforolimus trended toward improved PFS for bone tumors but did not achieve statistical significance (HR 0.70, 95% CI upper limit >1); the study was not powered for subgroup analyses16.\n\nEverolimus has also demonstrated some activity in osteosarcoma; in a pediatric phase I study, one of two enrolled osteosarcoma patients experienced prolonged stable disease for eight courses17. Everolimus has been shown to decrease drug-induced resistance to sorafenib via abrogation of the upregulation of mTORC1 and mTORC2 in murine models18; the combination of sorafenib and everolimus in an adult phase II study in patients with recurrent osteosarcoma with measurable disease yielded a 6-month PFS of 45%19. These results are in stark contrast to those observed in an historical cohort of osteosarcoma patients enrolled in phase II trials (Lagmay J et al., J Clin Oncol, in press). A recently completed phase I study of everolimus in combination with pazopanib for the treatment of adults with advanced solid tumors demonstrated tolerability; prolonged stable disease or partial response was observed for several patients with PI3K/AKT/mTOR pathway alterations20. Not all published studies have supported activity of mTOR inhibition in osteosarcoma; prior combinations of rapamycin with cyclophosphamide and temsirolimus with the insulin-like growth factor-1 receptor (IGF-1R) antibody cixutumumab failed to demonstrate efficacy in phase II studies including recurrent osteosarcoma patients21–23. However, when viewed as a whole, the available clinical data suggest potential activity of inhibition of the PI3K/mTOR pathway in osteosarcoma, which warrants further investigation.\n\n\nConclusions\n\nThe above data present compelling evidence of the role for dysregulation of the PI3K/mTOR pathway in osteosarcoma and suggest an opportunity for focused therapeutic strategies. Prior studies of mTOR inhibitors in osteosarcoma have demonstrated a hint of activity for targeting this pathway but were not adequate to assess the activity of PI3K/mTOR inhibition in osteosarcoma due to study design, lack of molecular correlation, and/or the number of patients enrolled. Furthermore, evidence that combinations of PI3K and mTOR inhibition can overcome patterns of resistance observed in single agent exposures and combination therapy shows potential promise for clinical activity. Given the paucity of active agents available for the treatment of recurrent and refractory osteosarcoma, inhibition of PI3K and mTOR may present a viable treatment strategy deserving of clinical investigation. It remains unclear whether the presence of aberrations within the pathway truly function as biomarkers of susceptibility to targeted agents; therefore, as a component of any future prospective studies of PI3K/mTOR inhibition, genomic analysis and assessment of activity of the pathway should be included as correlative biology studies.",
"appendix": "Competing interests\n\n\n\nThe authors have no conflicts of interest to disclose.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMirabello L, Troisi RJ, Savage SA: Osteosarcoma incidence and survival rates from 1973 to 2004: data from the Surveillance, Epidemiology, and End Results Program. Cancer. 2009; 115(7): 1531–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyers PA, Schwartz CL, Krailo M, et al.: Osteosarcoma: a randomized, prospective trial of the addition of ifosfamide and/or muramyl tripeptide to cisplatin, doxorubicin, and high-dose methotrexate. J Clin Oncol. 2005; 23(9): 2004–11. PubMed Abstract | Publisher Full Text\n\nKager L, Zoubek A, Pötschger U, et al.: Primary metastatic osteosarcoma: presentation and outcome of patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols. J Clin Oncol. 2003; 21(10): 2011–8. PubMed Abstract | Publisher Full Text\n\nEbb D, Meyers P, Grier H, et al.: Phase II trial of trastuzumab in combination with cytotoxic chemotherapy for treatment of metastatic osteosarcoma with human epidermal growth factor receptor 2 overexpression: a report from the children's oncology group. J Clin Oncol. 2012; 30(20): 2545–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBriccoli A, Rocca M, Salone M, et al.: High grade osteosarcoma of the extremities metastatic to the lung: long-term results in 323 patients treated combining surgery and chemotherapy, 1985–2005. Surg Oncol. 2010; 19(4): 193–9. PubMed Abstract | Publisher Full Text\n\nKempf-Bielack B, Bielack SS, Jürgens H, et al.: Osteosarcoma relapse after combined modality therapy: an analysis of unselected patients in the Cooperative Osteosarcoma Study Group (COSS). J Clin Oncol. 2005; 23(3): 559–68. PubMed Abstract | Publisher Full Text\n\nBielack SS, Kempf-Bielack B, Branscheid D, et al.: Second and subsequent recurrences of osteosarcoma: presentation, treatment, and outcomes of 249 consecutive cooperative osteosarcoma study group patients. J Clin Oncol. 2009; 27(4): 557–65. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFreeman SS, Allen SW, Ganti R, et al.: Copy number gains in EGFR and copy number losses in PTEN are common events in osteosarcoma tumors. Cancer. 2008; 113(6): 1453–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChoy E, Hornicek F, MacConaill L, et al.: High-throughput genotyping in osteosarcoma identifies multiple mutations in phosphoinositide-3-kinase and other oncogenes. Cancer. 2012; 118(11): 2905–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen X, Bahrami A, Pappo A, et al.: Recurrent somatic structural variations contribute to tumorigenesis in pediatric osteosarcoma. Cell Rep. 2014; 7(1): 104–12. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nPerry JA, Kiezun A, Tonzi P, et al.: Complementary genomic approaches highlight the PI3K/mTOR pathway as a common vulnerability in osteosarcoma. Proc Natl Acad Sci U S A. 2014; 111(51): E5564–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupte A, Baker EK, Wan SS, et al.: Systematic Screening Identifies Dual PI3K and mTOR Inhibition as a Conserved Therapeutic Vulnerability in Osteosarcoma. Clin Cancer Res. 2015; 21(14): 3216–29. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMoriarity BS, Otto GM, Rahrmann EP, et al.: A Sleeping Beauty forward genetic screen identifies new genes and pathways driving osteosarcoma development and metastasis. Nat Genet. 2015; 47(6): 615–24. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nPenel-Page M, Ray-Coquard I, Larcade J, et al.: Off-label use of targeted therapies in osteosarcomas: data from the French registry OUTC'S (Observatoire de l'Utilisation des Thérapies Ciblées dans les Sarcomes). BMC Cancer. 2015; 15: 854. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nChawla SP, Staddon AP, Baker LH, et al.: Phase II study of the mammalian target of rapamycin inhibitor ridaforolimus in patients with advanced bone and soft tissue sarcomas. J Clin Oncol. 2012; 30(1): 78–84. PubMed Abstract | Publisher Full Text\n\nDemetri GD, Chawla SP, Ray-Coquard I, et al.: Results of an international randomized phase III trial of the mammalian target of rapamycin inhibitor ridaforolimus versus placebo to control metastatic sarcomas in patients after benefit from prior chemotherapy. J Clin Oncol. 2013; 31(19): 2485–92. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFouladi M, Laningham F, Wu J, et al.: Phase I study of everolimus in pediatric patients with refractory solid tumors. J Clin Oncol. 2007; 25(30): 4806–12. PubMed Abstract | Publisher Full Text\n\nPignochino Y, Dell'Aglio C, Basirico M, et al.: The Combination of Sorafenib and Everolimus Abrogates mTORC1 and mTORC2 upregulation in osteosarcoma preclinical models. Clin Cancer Res. 2013; 19(8): 2117–31. PubMed Abstract | Publisher Full Text\n\nGrignani G, Palmerini E, Ferraresi V, et al.: Sorafenib and everolimus for patients with unresectable high-grade osteosarcoma progressing after standard treatment: a non-randomised phase 2 clinical trial. Lancet Oncol. 2015; 16(1): 98–107. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nRodrigues HV, Ke D, Lim J, et al.: Phase I combination of pazopanib and everolimus in PIK3CA mutation positive/PTEN loss patients with advanced solid tumors refractory to standard therapy. Invest New Drugs. 2015; 33(3): 700–9. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nSchwartz GK, Tap WD, Qin LX, et al.: Cixutumumab and temsirolimus for patients with bone and soft-tissue sarcoma: a multicentre, open-label, phase 2 trial. Lancet Oncol. 2013; 14(4): 371–82. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWagner LM, Fouladi M, Ahmed A, et al.: Phase II study of cixutumumab in combination with temsirolimus in pediatric patients and young adults with recurrent or refractory sarcoma: a report from the Children's Oncology Group. Pediatr Blood Cancer. 2015; 62(3): 440–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchuetze SM, Zhao L, Chugh R, et al.: Results of a phase II study of sirolimus and cyclophosphamide in patients with advanced sarcoma. Eur J Cancer. 2012; 48(9): 1347–53. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14769",
"date": "06 Jul 2016",
"name": "Erik H Danen",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14051",
"date": "06 Jul 2016",
"name": "Giovanni Grignani",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1590
|
https://f1000research.com/articles/5-1577/v1
|
05 Jul 16
|
{
"type": "Research Article",
"title": "Hospital and patient influencing factors of treatment schemes given to type 2 diabetes mellitus inpatients in Inner Mongolia, China",
"authors": [
"Nan Zhang",
"Edward McNeil",
"Sawitri Assanangkornchai",
"Yancun Fan",
"Nan Zhang",
"Edward McNeil"
],
"abstract": "Background: In clinical practice, the physician’s treatment decision making is influenced by many factors besides the patient’s clinical conditions and is the fundamental cause of healthcare inequity and discrimination in healthcare settings. Type 2 diabetes mellitus (T2DM) is a chronic disease with high prevalence, long average length of stay and high hospitalization rate. Although the treatment of T2DM is well guideline driven, there is a large body of evidence showing the existence of treatment disparities. More empirical studies from the provider side are needed to determine if non-clinical factors influence physician’s treatment choices.\n\nObjective: To determine the hospital and patient influencing factors of treatment schemes given to T2DM inpatients in Inner Mongolia, China.\n\nMethods: A cross-sectional, hospital-based survey using a cluster sampling technique was conducted in three tertiary hospitals and three county hospitals in Inner Mongolia, China. Treatment schemes were categorized as lifestyle management, oral therapy or insulin therapy according to the national guideline. Socio-demographic characteristics and variables related to severity of disease at the individual level and hospital level were collected. Weighted multinomial logistic regression models were used to determine influencing factors of treatment schemes.\n\nResults: Regardless of patients’ clinical conditions and health insurance types, both hospital and patient level variables were associated with treatment schemes. Males were more likely to be given oral therapy (RRR=1.72, 95% CI=1.06-2.81) and insulin therapy (RRR=1.94, 95% CI=1.29-2.91) compared to females who were given lifestyle management more frequently. Compared to the western region, hospitals in the central regions of Inner Mongolia were less likely to prescribe T2DM patients oral therapy (RRR = 0.18, 95% CI=0.05-0.61) and insulin therapy (RRR = 0.20, 95% CI=0.06-0.67) than lifestyle management. Compared with non-reformed tertiary hospitals, reformed tertiary hospitals and county hospitals were less likely to give T2DM patients oral therapy (RRR = 0.07 and 0.1 respectively) and insulin therapy (RRR = 0.11 and 0.17 respectively).\n\nConclusion: Gender was the only socio-demographic factors associated with treatment scheme for T2DM patients. Hospitals from different regions have different T2DM treatment patterns. Implementation of reform was shown to be associated with controlling medication use for T2DM inpatients. Further studies are needed to investigate the causes of unreasonable treatment disparities so that policies can be generated accordingly.",
"keywords": [
"Treatment scheme",
"healthcare disparities",
"T2DM",
"influence factors",
"hospital management"
],
"content": "Introduction\n\nProviding equitable and high quality healthcare is an essential objective of the health systems all over the world. However, variations in medical practices are observed across and within countries1,2 in the management of many diseases and clinical conditions3,4, including diagnosis, nursing care and treatment5.\n\nThe differences in healthcare among different population groups are defined as healthcare disparities6. More and more attention is now focused on health disparities that impair healthcare quality and overall health outcomes. It is estimated that $309 billion per year is lost due to the direct and indirect costs of health disparities in the United States of America7.\n\nBoth clinical and non-clinical factors may lead to disparities in healthcare. From the view of healthcare equity, it is important to recognize the wide range of factors that contribute to disparities, especially non-clinical factors, such as socio-economic and healthcare system factors. Evidence from the population side suggests that a variety of factors are related to unequal health services received by people8–10. Although the influencing factors differ among countries and regions, demographic and socio-economic factors are proposed in many studies11,12. Since the clinical conditions of the people are very difficult to evaluate from most population-based surveys by self-reported health status, there are biases when comparisons of treatments are made between different population groups. A hospital-based study may overcome this limitation and provide physicians defined clinical and non-clinical patient information, so that they have more confidence in comparing the healthcare services they receive. Yet evidence from the provider side is limited, especially in low- and middle-income counties13.\n\nInner Mongolia is a self-governed province located in the northern part of China. Compared to other parts of China, healthcare accessibility is inferior in Inner Mongolia because of the low population density, limited health resources and underdeveloped social-economic status. All of these characteristics may promote healthcare disparities.\n\nDiabetes has become a large and rapidly-growing health problem. In 2011, it accounted for 8.2% of global all-cause mortality of people aged 20–7914. Type 2 diabetes mellitus (T2DM) is one of the leading chronic diseases in Inner Mongolia. The prevalence of diabetes in Inner Mongolia was 2.6% in 2008, according to the fourth National Health Service survey, an increase of 160% from 200315. Besides having a high prevalence, diabetes is also associated with longer length of stay (LOS) in hospital and incurs a high medical expenditure. According to the data of Health Ministry of China in the year 2010 the average LOS of diabetes inpatients was 13.2 days, which was 3.2 days longer than the average LOS of all inpatients16. Healthcare disparities may exist in treatment of T2DM inpatients since the disease is chronic and more expensive than other diseases.\n\nTreatment of T2DM is well guideline-driven. The Chinese Diabetes Society (CDS) issued the Chinese Guideline for Treatment of T2DM and has revised it every 3 to 4 years17. The International Diabetes Federation (IDF) and the American Diabetes Association (ADA) provide updated guidelines for T2DM treatment as well18. Treatment schemes in all the T2DM treatment guidelines consistently include three components: lifestyle management, oral therapy and insulin therapy17,19. Since type 2 diabetes is a progressive disease, patients who have had a longer duration of disease are likely to have reduced beta-cell function and require more intensive therapy compared to patients with a more recently diagnosed disease20.\n\nA hospital-based study was implemented aiming to determine the hospital and patient influencing factors on treatment schemes given to T2DM inpatients in Inner Mongolia of China. Both clinical and non-clinical factors from hospital and patient perspectives were considered. The potential influencing factors considered in this study were categorized as socio-demographic (non-clinical factors in patient side), disease characteristics (clinical factors in hospital side) and hospital characteristics (non-clinical factors in hospital side).\n\n\nMethods\n\nA cross-sectional hospital-based study.\n\nThe population of this study comprised T2DM inpatients in Inner Mongolia of China. Inner Mongolia has three-tier hospitals. Tertiary hospitals located in urban area serve both urban and rural people. County hospitals are secondary hospitals serving rural people mostly. Community health institutions and township hospitals provide primary healthcare in urban and rural areas respectively. Moreover, there are Mongolian hospitals included in the hospital system of Inner Mongolia that mainly serve minority people and use totally different techniques and drugs from western medicine. Tertiary and county hospitals, as major providers of T2DM hospitalization care in Inner Mongolia, were chosen for this study. A multistage cluster sampling method was used to select the study sample. Inner Mongolia was geographically classified into three regions and the largest tertiary hospital and county hospital in each region were purposively selected. The eligible hospitals should have facilities to provide standard T2DM inpatient services as required by clinical guidelines. Only hospitals with good quality medical records and those where hospital charge information could be extracted from hospital information system were selected. Three tertiary hospitals and three county hospitals were finally chosen. All consecutive inpatients with a principle diagnosis of T2DM (ICD-10 codes: E11.2–E11.9) admitted into the sampled hospitals and discharged during the data collection period were recruited into this study. Those who stayed in hospital for less than 24 hours or could not communicate by themselves were excluded. Finally, a total of 771 eligible participants were recruited into this study.\n\nTable 1 shows the characteristics of surveyed hospitals.\n\n* Data were obtained from yearly statistical data of each hospital.\n\n** Data were extracted from the hospital information system (HIS) of each hospital.\n\nOutcome variables. The outcome variable of this study was treatment schemes of T2DM inpatients, which were classified into three categories: lifestyle management, oral therapy and insulin therapy, according to China’s treatment guideline for T2DM.\n\nExplanatory variables. Socio-demographic variables (age, gender, ethnicity, marital status, occupation, residence, education, yearly income and expenditure, insurance scheme), disease characteristics variables (duration of T2DM, Age-adjusted Charlson Comorbidity Index (ACCI) score, diagnosis, participation in treatment planning), and institutional variables (hospital level, location of hospital and hospital reform status) were considered as explanatory variables for treatment schemes of T2DM inpatients.\n\n\nData collection and measurement\n\nPatient interview data and medical records were collected and analyzed. Patient interviews were conducted one day before the patients were discharged using a questionnaire administered by well-trained research assistants. One week after the interview, the patients’ final medical records were reviewed by two researchers using a self-designed medical record review form. Both the patient interview questionnaire and the medical review forms were pre-tested in a pilot study for validation. A standard operating procedure was generated to standardize the data collection process.\n\nSocio-demographic variables and two clinical variables (duration of T2DM and participation in treatment planning) were collected by patient interview, the other variables were extracted from the medical records.\n\nDuration of T2DM was calculated by asking a patient the exact year when he/she was first diagnosed as T2DM by a clinician. ACCI scores were calculated automatically using an online calculator, with scores ranging from 0–37. The calculation was done by adding the weighted scores for 19 medical conditions. High scores are associated with a poorer prognosis21. Diagnosis was coded as T2DM, T2DM with complications, T2DM with comorbidities, and T2DM with complications and comorbidities. Patient participation in treatment planning was determined by asking patients: “Did the doctors discuss your treatment plan with you before they made a decision?” Hospital reform status was confirmed with the local government. Only the hospital which had implemented the reform policies (including Zero mark-up on part of essential medicines and clinical pathway for T2DM) was defined as reformed hospital. Two of the surveyed county hospitals were assigned as pilot hospital of hospital reform under the framework of national medical reform, however, both of them were coded as non-reformed hospitals for having not started reform yet.\n\n\nStatistical methods\n\nData were presented as percentages for categorical variables or means for continuous variables. Multinomial logistic regression was used to explore the factors associated with treatment schemes. Weights were used to adjust the estimates due to cluster sampling. The strength of the association between factors and treatment scheme were presented as relative risk ratios (RRR) and 95% confidence intervals (95% CI). A P-value of less than 0.05 was considered as statistically significant. Data analysis was performed using the survey package of R software version 3.0.3 and Stata version 10.0.\n\n\nEthical considerations\n\nThis study was approved by the Institutional Review Board of the Faculty of Medicine, Prince of Songkla University, Thailand. Permission was obtained from surveyed hospitals and written informed consent was obtained from each participant before enrollment in the study. No incentives or financial payments were provided for the interviewed patients. Personal identification was removed from the completed questionnaires and confidentiality was assured.\n\n\nResults\n\nSocial-demographic characteristics of participants and their treatment schemes are shown in Table 2. 80% of T2DM inpatients were aged 50 years and above. 6.4% attained a bachelor degree study or above. Minority groups were rare (3.7%). More than half of the participants (59.5%) were living in urban areas. Although the national health insurance system (urban employee’s insurance scheme, urban resident’s insurance scheme and new rural cooperative medical system) covered most of the participants, 21.2% of the participants covered the cost of their hospitalization.\n\nUEIS: Urban employee’s insurance scheme\n\nURIS: Urban resident’s insurance scheme\n\nNRCMS: New rural cooperative medical system\n\nTreatment schemes significantly differed between genders. Females were more likely to be given lifestyle management and oral therapy, whereas males were more likely to be given insulin therapy. Lifestyle management and oral therapy were mostly given to those aged 60–69, those aged 50–59 were more likely to be given insulin therapy, however, the p-value was just on the cut point of statistical significance.\n\nTable 3 shows treatment schemes in relation to T2DM inpatients’ clinical characteristics. Diagnosis was significantly associated with treatment schemes. Those with comorbidities were more likely to be given lifestyle management and oral therapy than those with other clinical types, whereas those with complications and comorbidities were more likely to be given insulin therapy. The average duration of T2DM was 7.3 years (SE=0.5) and the mean ACCI score was 5.0 (SE=0.5). Physicians discussed with most of T2DM inpatients (85%) when they planned the treatment scheme for them.\n\nTable 4 shows the T2DM inpatient treatment schemes in different hospitals. Treatment schemes for T2DM inpatients differed in hospitals located in different regions. Lifestyle management and oral therapy were given more in the Eastern region, whereas insulin therapy was given more in the Western and central regions. Since there were no county hospitals served as pilot hospital for medical reform when we conducted this survey, hospital level and reform status were combined into one category variable to be analyzed. In the tertiary hospitals, which were not involved in medical reform, insulin therapy was common, whereas in the tertiary hospital with reform, oral therapy was used most. County hospital gave T2DM inpatients lifestyle management more frequently than medication. However, all the differences above were not statistically significant.\n\nTable 5 summarizes the results of a multivariate analysis for hospital and patient factors associated with treatment schemes given to T2DM inpatients. After controlling for patients’ clinical characteristics and payers, both hospital and patient level variables were associated with treatment schemes. Males were more likely to be given oral therapy (RRR = 1.72) and insulin therapy (RRR = 1.94) than females. Compared to other clinical types, those with comorbidities were less likely to be given insulin therapy (RRR = 0.16). Compared with the Western region, hospitals in central regions of Inner Mongolia were less likely to give T2DM inpatients oral therapy (RRR = 0.18) and insulin therapy (RRR = 0.20) than lifestyle management. Compared with tertiary hospitals without doing any healthcare reform, tertiary hospitals with reform and county hospitals were less likely to give T2DM inpatients oral therapy (RRR = 0.07 and 0.1 respectively) and insulin therapy (RRR = 0.11 and 0.17 respectively).\n\nUEIS: Urban employee’s insurance scheme\n\nURIS: Urban resident’s insurance scheme\n\nNRCMS: New rural cooperative medical system\n\nRRR: relative risk ratio. Reference outcome group = lifestyle management\n\nCI: confidence interval\n\n\nDiscussion\n\nAfter controlling for clinical conditions such as duration of disease, ACCI score and diagnosis, the hospital level factor stood out of all the potential explanatory variables. Hospital location, reform status and hospital level were associated with treatment scheme for T2DM inpatients significantly, while gender was the only demographic variable related.\n\nCompared with the Western region, T2DM patients treated in the central regions of Inner Mongolia were more likely to be given lifestyle management than medication. As one of the major causes for healthcare disparities, geographic variations in treatment pattern of T2DM have been reported in other studies22. However, the underlying reasons of the variations were not always the same23. There might be two possible explanations for regional variations in this study. First, the Western region in Inner Mongolia has a lower population density compared with the central region (23 person/km2 vs 174 person/km2). Since a general practice system has not been fully set up in China, especially in Inner Mongolia, inferior accessibility to healthcare may reduce the healthcare seeking behavior of T2DM patients, which will cause them to lose the opportunity for early treatment. Second, hospitals in the same region have more chance to have professional and academic exchanges and learn from each other under the circumstance of lacking clinical pathway management. More evidences are needed to get the precise information on the factors contributing to the variations so that policies could be designed accordingly.\n\nIn this study we show that implementation of reform was associated with controlling medication use for T2DM patients. Reformed tertiary hospitals were less likely to give T2DM patients medication treatment compared to tertiary hospitals without any reform. The public hospital reform in China under the national healthcare reform framework started from tertiary hospitals24. In the first stage of the reform, enhancing internal management was carried out by many hospitals to improve the quality of care. Zero mark-up on part of essential medicines (A zero-profit drug policy introduced by China government from the year 2009 to remove the mark-up for medicines sale in hospitals. The major objective of this policy was to reduce the incentive for providers to prescribe unnecessary drugs25.), and clinical pathway for T2DM were implemented in the reformed hospitals of this study. Although the hospitals had set out to implement the reform for less than three years when this survey was conducted, a series of policies was generalized and experimented by the hospitals to match the requirement of the government. Some studies report that the zero mark-up policy can decrease drug prescriptions and reduce total expense for both outpatient and inpatient services26,27. Clinical pathway has been shown to reduce the variability in clinical practices28. However, further studies are needed to confirm the causation between the reform and the change in physicians’ clinical practices.\n\nAmong non-clinical characteristics, major gender differences in the treatment schemes given to T2DM patients were observed in this study. Males were more likely to be given oral therapy and insulin therapy than females who were mostly given lifestyle management. Many studies have revealed the fact that females have increased adherence to preventive practices29, and males have lower obedience to diet control and exercise30. However, gender differences in treatment protocols lead to unequal treatment and it should be avoided as much as possible. One way this can be done is by incorporating patient preferences into the treatment schemes. Systematic reviews done on the patient preferences for non-insulin diabetes medications during 2007–201231,32 have shown the importance of incorporating patient preferences into treatment decisions. These reviews showed that the key attributes of diabetes medication associated with patient preferences include treatment benefits (e.g. glycemic control and weight loss/control), treatment burden (e.g. administration, frequency, and cost), and side effects (e.g. weight gain, gastrointestinal effects, and hypoglycemia)31. Therefore, gender differences in the treatment schemes can be allowed as long as they accommodate the patient’s preferences and treatment efficacy is not compromised.\n\nA number of studies have found evidence about health disparities in different ethnic groups33,34. However, ethnicity was not significantly associated with treatment schemes in this study. If we look at the distribution of ethnicity among T2DM inpatients, it was quite different from that of the whole population of Inner Mongolia. Our findings showed that minority groups with T2DM were rare (3.7%) across the six hospitals studied. This proportion was much lower than in the whole population of Inner Mongolia (21%). This could be because minorities prefer traditional medicine to Western medicine. However, further studies are needed to obtain information about accessibility and utilization of healthcare for minority groups with T2DM in Inner Mongolia. Where ethnic and racial disparities exist in the treatment of diabetes patients, it is evident that policy responses for addressing minority groups are needed.\n\nEducation level was low among T2DM patients in this study. Only 6.5% of T2DM patients had attained a bachelor degree or above, which was lower than the whole population (10.3%). Education has been found to be associated with all aspects of T2DM treatment including prevention, outpatient care, inpatient care, and rehabilitation29,35. However, other studies have put forward evidence that the educational level of patients had no association with untreated diabetes36 or with the prevalence of diabetes37. Although the possibilities for disparities in treatment types may exist by level of education, this study did not find any association between education level and the type of treatment schemes that was given to the patients.\n\nThis study also found that although the national health insurance system covered most (77.6%) of the participants, 21.2% of the participants had no insurance. In contrast, among all adults with diabetes in the USA, 92.0% had some form of health insurance38. Physicians usually have different personal attitudes toward insured and uninsured patients. This was demonstrated in an analysis of patients who were insured which showed that they had a higher number of prescribed medications, and a higher total price of prescription than those who paid cash only39. It has been found that even insurers have major differences in attitude with regard to diabetes. Some patients who had declared their disease status to the insurers were refused acceptance for insurance and for some, their premiums were increased40. These evidences prove the importance of state mandated insurance programs, especially for diseases such as T2DM which has a long treatment period.\n\n\nStrengths and limitations\n\nThis was a hospital-based study using combined methods of patient interview and hospital record review to obtain clinical and non-clinical data, which can eliminate the recall bias on health service experience inherent in population based surveys. However, several limitations in this study should be acknowledged. First, it was a hospital-based study, therefore the findings cannot be generalized to the whole population of China, or T2DM patients. Second, due to the cross-sectional nature of the study, causality of certain risk factors and the study outcome cannot be established. Third, we did not measure all potential patient risk factors, such as hemoglobin A1C values, which were not available. As such, duration of T2DM, diagnosis and ACCI score were used to estimate severity of disease, which are known to be less precise. Fourth, physicians’ characteristics (such as age, how long they were certified, etc.) were not available, which would be a factor in determining treatment and adherence to the national guidelines.\n\n\nConclusion\n\nIn conclusion, this study did not show any socio-demographic factors except gender influencing the treatment patterns for T2DM inpatients, independently of their clinical condition. Hospitals from different regions have different treatment patterns, which tends to increase the healthcare disparities and should be eliminated by exploring further policy strategies. Implementation of reform was shown to be associated with controlling medication use for T2DM inpatients. However, the causation between the reform and changing of physicians’ clinical practices should be explored in further studies.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data of hospital and patient influencing factors on treatment schemes given for type 2 diabetes mellitus in patients in Inner Mongolia, 10.5256/f1000research.9095.d12782541",
"appendix": "Author contributions\n\n\n\nNZ, SA and YF were principal investigators of this study. NZ and SA developed the concept and study design. NZ and YF collected the data. NZ wrote the manuscript. SA and EM assisted in development of data analysis, supervised the study, and revised this manuscript. SA and YF provided consultations for drafting this manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no financial, personal, or professional competing interests to declare in connection with this article and no conflicts of interest in the authorship or publication of this contribution.\n\n\nGrant information\n\nThe study was funded by the China Medical Board through the project titled “Joint Research Capacity Strengthening of the Western Rural Health Network, China” under Prince of Songkla University and a grant from the Graduate School, Prince of Songkla University, Thailand, and the Key Research Institute of Humanities & Social Science in Inner Mongolia, China.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThis study is part of the first author’s thesis in partial fulfillment of the requirements for a Ph.D. at Prince of Songkla University, Thailand, and the China Medical Board under the project of “Join Research Capacity Strengthening of the Western Rural Health Network, China” provided the financial support. The authors thank all participants for data collection in this study.\n\n\nReferences\n\nRubin RR, Peyrot M, Siminerio LM: Health care and patient-reported outcomes: results of the cross-national Diabetes Attitudes, Wishes and Needs (DAWN) study. Diabetes Care. 2006; 29(6): 1249–1255. PubMed Abstract | Publisher Full Text\n\nMillett C, Gray J, Saxena S, et al.: Ethnic disparities in diabetes management and pay-for-performance in the UK: the Wandsworth Prospective Diabetes Study. PLoS Med. 2007; 4(6): e191. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPopescu I, Vaughan-Sarrazin MS, Rosenthal GE: Differences in mortality and use of revascularization in black and white patients with acute MI admitted to hospitals with and without revascularization services. JAMA. 2007; 297(22): 2489–2495. PubMed Abstract | Publisher Full Text\n\nKelly DL, Dixon LB, Kreyenbuhl JA, et al.: Clozapine utilization and outcomes by race in a public mental health system: 1994–2000. J Clin Psychiatry. 2006; 67(9): 1404–1411. PubMed Abstract | Publisher Full Text\n\nMillett C, Car J, Eldred D, et al.: Diabetes prevalence, process of care and outcomes in relation to practice size, caseload and deprivation: national cross-sectional study in primary care. J R Soc Med. 2007; 100(6): 275–283. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHunt JB, Eisenberg D, Lu L, et al.: Racial/Ethnic Disparities in Mental Health Care Utilization among U.S. College Students: Applying the Institution of Medicine Definition of Health Care Disparities. Acad Psychiatry. 2015; 39(5): 520–6. PubMed Abstract | Publisher Full Text\n\nLaVeist TA, Gaskin D, Richard P: Estimating the economic burden of racial health inequalities in the United States. Int J Health Serv. 2011; 41(2): 231–238. PubMed Abstract | Publisher Full Text\n\nWang KW, Shu ZK, Cai L, et al.: Assessment of the magnitude of contextual and individual demographic effects on diabetes mellitus and glucose intolerance in rural Southwest China: a multilevel analysis. PLoS One. 2013; 8(7): e68553. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Z, Su Y, Gao J, et al.: Assessing equity of healthcare utilization in rural China: results from nationally representative surveys from 1993 to 2008. Int J Equity Health. 2013; 12: 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Z, Gao J, Fox A, et al.: Measuring the equity of inpatient utilization in Chinese rural areas. BMC Health Serv Res. 2011; 11: 201. PubMed Abstract | Publisher Full Text | Free Full Text\n\nErgin I, Mandiracioglu A: Demographic and socioeconomic inequalities for self-rated health and happiness in elderly: the situation for Turkey regarding World Values Survey between 1990 and 2013. Arch Gerontol Geriatr. 2015; 61(2): 224–30. PubMed Abstract | Publisher Full Text\n\nSmith NR, Lewis DJ, Fahy A, et al.: Individual socio-demographic factors and perceptions of the environment as determinants of inequalities in adolescent physical and psychological health: the Olympic Regeneration in East London (ORiEL) study. BMC Public Health. 2015; 15: 150. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgardh E, Allebeck P, Hallqvist J, et al.: Type 2 diabetes incidence and socio-economic position: a systematic review and meta-analysis. Int J Epidemiol. 2011; 40(3): 804-818. PubMed Abstract | Publisher Full Text\n\nFederation ID: IDF Diabetes Atlas. 2011; 5th edition. Reference Source\n\nMongolia HDol: The Fourth National Health Service survey-Report of Inner Mongolia. Hohhot, Inner Mongolia. 2010.\n\nChina NHaFPCotPsRo: National Health Year Book 2011. Statistic of inpatients in the year 2010. 2011. Reference Source\n\nSociety CD: Chinese guideline for the management of type 2 diabetes mellitus (2013 edition). Chin J Diabetes Mellitus. 2014; 6: 447–498.\n\nAssoc AD: Standards of Medical Care in Diabetes--2014. Diabetes Care. 2014; 37(Suppl 1): S14–S80. PubMed Abstract | Publisher Full Text\n\nInzucchi SE, Matthews DRManagement of Hyperglycemia in Type 2 Diabetes American Diabetes Association and European Association for the Study of Diabetes Position Statement Writing Group:Response to Comments on Inzucchi et al. Management of Hyperglycemia in Type 2 Diabetes, 2015: A Patient-Centered Approach. Update to a Position Statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 2015;38:140–149. Diabetes Care. 2015; 38(8): E128–E129. PubMed Abstract | Publisher Full Text\n\nJi LN, Lu JM, Guo XH, et al.: Glycemic control among patients in China with type 2 diabetes mellitus receiving oral drugs or injectables. BMC Public Health. 2013; 13: 602. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang CC, Chen PC, Hsu CW, et al.: Validity of the age-adjusted Charlson comorbidity index on clinical outcomes for patients with nasopharyngeal cancer post radiation treatment: a 5-year nationwide cohort study. PLoS One. 2015; 10(1): e0117323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTamayo T, Claessen H, Rückert IM, et al.: Treatment Pattern of Type 2 Diabetes Differs in Two German Regions and with Patients' Socioeconomic Position. Plos One. 2014; 9(6): e99773. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZuckerman S, Waidmann T, Berenson R, et al.: Clarifying Sources of Geographic Differences in Medicare Spending. N Engl J Med. 2010; 363(1): 54–62. PubMed Abstract | Publisher Full Text\n\nShanlian Hu CL, Public Hospital Reform: Sweden: Swedish Agency for Growth Policy Analysis.2013.\n\nWenhui M, Wen C: The Zero Mark-up Policy for essential medicines at primary level facilities - China case study. WHO, 2015. Reference Source\n\nZhou ZL, Su YF, Campbell B, et al.: The Financial Impact of the 'Zero-Markup Policy for Essential Drugs' on Patients in County Hospitals in Western Rural China. PLoS One. 2015; 10(3): e0121630. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang H, Hu H, Wu C, et al.: Impact of China's Public Hospital Reform on Healthcare Expenditures and Utilization: A Case Study in ZJ Province. PLoS One. 2015; 10(11): e0143130. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPanella M, Marchisio S, Di Stanislao F: Reducing clinical variations with clinical pathways: do pathways work? Int J Qual Health Care. 2003; 15(6): 509–521. PubMed Abstract | Publisher Full Text\n\nJimenez-Trujillo I, Jimenez-Garcia R, Esteban-Hernandez J, et al.: Predictors of Adherence to Multiple Clinical Preventive Recommendations among Adults with Diabetes in Spain. PLoS One. 2015; 10(6): e0131844. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi N, Yang XF, Deng Y, et al.: [Diabetes self-management and its association with diabetic retinopathy in patients with type 2 diabetes]. Zhonghua Yan Ke Za Zhi. 2013; 49(6): 500–506. PubMed Abstract | Publisher Full Text\n\nPurnell TS, Joy S, Little E, et al.: Patient preferences for noninsulin diabetes medications: a systematic review. Diabetes Care. 2014; 37(7): 2055–2062. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpaan N, Teplova A, Stam G, et al.: Systematic review: continuous intraperitoneal insulin infusion with implantable insulin pumps for diabetes mellitus. Acta Diabetol. 2014; 51(3): 339–351. PubMed Abstract | Publisher Full Text\n\nWagner DV, Stoeckel M, E Tudor M, et al.: Treating the most vulnerable and costly in diabetes. Curr Diab Rep. 2015; 15(6): 606. PubMed Abstract | Publisher Full Text\n\nParrinello CM, Rastegar I, Godino JG, et al.: Prevalence of and Racial Disparities in Risk Factor Control in Older Adults With Diabetes: The Atherosclerosis Risk in Communities Study. Diabetes Care. 2015; 38(7): 1290–1298. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRubin RR, Peyrot M, Saudek CD: Effect of diabetes education on self-care, metabolic control, and emotional well-being. Diabetes Care. 1989; 12(10): 673–679. PubMed Abstract | Publisher Full Text\n\nBeltrán-Sánchez H, Drumond-Andrade FC, Riosmena F: Contribution of socioeconomic factors and health care access to the awareness and treatment of diabetes and hypertension among older Mexican adults. Salud Publica Mex. 2015; 57(Suppl 1): s06–14. PubMed Abstract | Free Full Text\n\nDa-Mata FA, Galvao TF, Pereira MG, et al.: Prevalence of Self-Reported Diabetes and Its Associated Factors: A Population-Based Study in Brazil. Int J Endocrinol. 2015; 2015: 610790. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris MI, Cowie CC, Eastman R: Health-insurance coverage for adults with diabetes in the U.S. population. Diabetes Care. 1994; 17(6): 585–591. PubMed Abstract | Publisher Full Text\n\nAl-Mohamadi A, Al-Harbi AM, Manshi AM, et al.: Medications prescribing pattern toward insured patients. Saudi Pharm J. 2014; 22(1): 27–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrier BM, Sullivan FM, Stewart EJ: Diabetes and insurance: a survey of patient experience. Diabet Med. 1984; 1(2): 127–130. PubMed Abstract | Publisher Full Text\n\nZhang N, Fan Y, McNeil E, et al.: Dataset 1 in: Hospital and patient influencing factors of treatment schemes given to type 2 diabetes mellitus inpatients in Inner Mongolia, China. F1000Research. 2016. Data Source"
}
|
[
{
"id": "14808",
"date": "11 Jul 2016",
"name": "Stephen J Nicholas",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper briefly sets out on page 3 the nature of equity issues in hospital health care and the problems of measuring health care disparities, especially from the clinical side. Using a transitional and developing region of China, Inner Mongolia, the paper focuses on the widely accepted practice of measuring disparities from the population side (socio-economic characteristics) factors. Importantly, the paper also focuses on different types of hospital and also measures clinical treatment differences for diabetes (T2DM) treatments. The choice of disease, T2DM, is justified as the research focus. The authors should be commended for their access to treatment records.\n\nOn page 3, the authors might add a couple of sentences on tertiary versus county hospitals, especially given the differences in bed-size. Also, the choice of hospital depended on the quality of the records, and a sentence on differences (if any) in tertiary versus county hospital records would be useful.\nIn terms of the explanatory variables (page 3), the authors might briefly define insurance scheme, hospital reform status and diagnosis in this section (although these are (partly) covered in the section on ‘Data collection and measurement’).\nOn page 8, the argument should make clear that type of insurance and ‘no insurance’ did not have an impact on treatment. The discussion contrasts the Inner Mongolia case of no differences in treat by insurance/’no insurance’ with other countries, but does not emphasize that there are no differences in the Inner Mongolia data between insurance/'no-insurance'. This requires only one sentence.",
"responses": []
},
{
"id": "14809",
"date": "12 Jul 2016",
"name": "Jian Wang",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a policy-relevant article to study hospital and patient factors influencing T2DM treatment in Mongolia. The rapid increasing of T2DM gives a pressure to the health system in Inner Mongolia. The authors are interested in exploring health care disparities by conducting a cross-section hospital survey in 6 hospitals in 3 regions. I would approve the article. However, there may be minor revisions if the authors accept the following comments.\nIn the limitation section, it would be better to explain if there any recall bias when using data from patient interview? If so, please state in limitation part.\n\nGive a few words to demonstrate “a series of policies was generalized and experimented by the hospitals” at the current health reform.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1577
|
https://f1000research.com/articles/5-900/v1
|
17 May 16
|
{
"type": "Case Report",
"title": "Case Report: Whole exome sequencing reveals a novel frameshift deletion mutation p.G2254fs in COL7A1 associated with autosomal recessive dystrophic epidermolysis bullosa",
"authors": [
"Shamsudheen Karuthedath Vellarikkal",
"Rijith Jayarajan",
"Ankit Verma",
"Sreelata Nair",
"Rowmika Ravi",
"Vigneshwar Senthivel",
"Sridhar Sivasubbu",
"Vinod Scaria",
"Shamsudheen Karuthedath Vellarikkal",
"Rijith Jayarajan",
"Ankit Verma",
"Sreelata Nair",
"Rowmika Ravi",
"Vigneshwar Senthivel"
],
"abstract": "Dystrophic epidermolysis bullosa simplex (DEB) is a phenotypically diverse inherited skin fragility disorder. It is majorly manifested by appearance of epidermal bullae upon friction caused either by physical or environmental trauma. The phenotypic manifestations also include appearance of milia, scarring all over the body and nail dystrophy. DEB can be inherited in a recessive or dominant form and the recessive form of DEB (RDEB) is more severe. In the present study, we identify a novel p.G2254fs mutation in COL7A1 gene causing a sporadic case of RDEB by whole exome sequencing (WES). Apart from adding a novel frameshift Collagen VII mutation to the repertoire of known mutations reported in the disease, to the best of our knowledge, this is the first report of a genetically characterized case of DEB from India.",
"keywords": [
"Dystrophic epidermolysis bullosa",
"simplex whole exome sequencing",
"Collagen VII mutation"
],
"content": "Introduction\n\nDystrophic epidermolysis bullosa (DEB) is an extremely rare subtype of epidermolysis bullosa with an estimated incidence of approximately 6.5 per million newborns. The disease is caused by mutations in collagen VII (COL7A1)1. Collagen VII is a major structural macromolecule of the skin and plays an important component of the anchoring fibrils which connect the epidermis and dermis of the skin. The disease affects the skin, the mucosa (including that of the oral cavity) and gastrointestinal tract. The blisters are further followed by scarring and development of deformities. The disease also predisposes individuals to development of skin cancer and it is estimated that almost all affected members develop cancers in the third or fourth decade of life.\n\n\nCase Report\n\nA 4.5-year-old South Indian female child presented to the outpatient clinic with a history of multiple vesicular and bullous lesions induced by trauma since perinatal period. The child was born out of a third degree consanguineous marriage with no known history of similar illness. The child had severe blistering and scarring all over the body, nail dystrophy and milia. The oral mucosa was involved along with tongue blistering, dental calculus, and chipping of teeth with difficulty in opening the mouth. The child also had flexural deformities resulting in contractures and pseudo-syndactyly of the fingers. The clinical picture (Figure 1a,b) corroborated the diagnosis of dystrophic epidermolysis bullosa (DEB). There is no center in India offering genetic diagnosis for the disease using targeted gene sequencing. Given that targeted gene sequencing can be quite expensive, tedious and time-consuming to standardise, we attempted whole-exome sequencing (WES). Moreover no background genetic map of mutations in the disease from India was available. Previous reports, including from our laboratory suggest WES as an alternative to traditional approaches; WES being fast, less tedious, and cost-effective while also providing a holistic view of the mutation spectrum of the patient2–4.\n\nHands and thoracic region showing generalized bullae, scarring and milia b) Lower legs showing scarring, bullae, milia and characteristic dystrophic nails c) Pedigree of the family d) The chromatogram depicting capillary sequencing results of c.6759_6760del in the trio. The mutation loci (ΔCT) is highlighted with asterisks e) Domain structure of COL7A1 protein showing Von Willebrand factor type A domain (VWA), Fibronectin type III domain (fn3), collagen triple helix domain (blue) and Kunitz domain (yellow). Each needle represents disease causing variation site and the red needle represent p.G2254fs (c.6759_6760del) variation. Panel at the bottom represents COL7A1 p.G2254fs induced PTC compared to the normal protein.\n\nApproximately 5 ml of blood was collected from the affected individual and the parents after obtaining signed informed consent and approval from the institutional ethical committee (BSC0212 IHECC proposal No.08). Genomic DNA was isolated by using salting out method5. 50ng of high quality DNA was used for whole exome sample preparation using a Nextera (Illumina Inc, USA) expanded exome kit according to manufacturer supplied instruction. The exome was sequenced by Illumina Hiseq2500 according to the manufacturer’s protocols (Illumina Inc, USA). Paired-end reads of 150 bases were generated, which was quality and adapter trimmed at a Phred quality score of 20. Alignment was performed on the human reference genome (hg19) using Burrows-Wheeler Alignment (version 0.5.10-evan.9)6. The mean mapped coverage on target region was 12.2x. Variants were called using Platypus pipelines (version 0.7.9.1)7. Analysis revealed a novel homozygous frameshift deletion (chr3:g.48610366CT>-) c.6759_6760del (p.G2254fs) in COL7A1 gene. The c.6759_6760del was predicted to be deleterious (confidence score 0.858) and introduce a premature termination codon (PTC) at 2273th amino acid position according to SIFT8. Homozygous PTCs in COL7A1 is previously reported to reduce overall stability of anchoring filaments and cause mild to very severe generalised RDEB1. Secondary structure analysis shows that p.G2254fs resultant PTC leads to loss of function of several collagen triple helix repeats and kunitz domain (Figure 1e). We also found a homozygous nonsynonymous variation c.5716C>T (p.P1906S) in COL7A1 which was predicted to be ‘tolerated’ by SIFT (0.5)9.\n\nThe variant was verified independently using capillary sequencing in the child and parents. The variant was not found in ExAC or our internal cohort of 122 exomes, confirming its rarity and novelty. Parents were provided detailed genetic counselling by the consulting clinical geneticist.\n\n\nDiscussion\n\nDystrophic EB could be inherited in both recessive and dominant form1. Several cases of DEB have been reported from India. A recent paper reported a cohort of 17 DEB patients using immunofluorescence mapping10, though the patients were not genetically characterized. Our earlier report characterized a novel mutation in KRT5 associated with epidermolysis bullosa (EB) simplex in West India4. Taken together, we suggest a large and potentially uncharacterized repertoire of genetic variations causing EB in India, which might benefit from genetic screening approaches.\n\nIn this study, we show the application of next-generation sequencing to identify the mutation in a sporadic case of autosomal recessive EB in clinical settings. Apart from adding a novel frameshift collagen VII deletion mutation to the repertoire of known mutations in the disease, to the best of our knowledge, this is the first report of a genetically characterized patient of DEB from India. We suggest that next-generation sequencing approach would significantly benefit the understanding and genetic characterization of this rare disease in India.\n\n\nConsent\n\nWritten informed consent was obtained from the parent of the patients for publication of this case report and any accompanying images and/or other details that could potentially reveal the patient’s identity.\n\n\nData availability\n\nThe raw exome sequencing data are available at the NCBI Sequence Read Archive (http://www.ncbi.nlm.nih.gov/sra), accession number SRX1584466.",
"appendix": "Author contributions\n\n\n\nSN performed the clinical evaluation and sent the blood for DNA analyses and provided genetic counselling. SKV, RJ, RR, AV and VSV isolated the DNA, conducted the quality checks, prepared the exome capture and sequencing library, performed the exome sequencing, performed data quality checks on the reads, reference alignments, variant call and computational prioritization of the variants, designed and performed the validation experiments and contributed to writing the manuscript. SS and VS conceptualized and oversaw the all the experiments and analysis and contributed to writing the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nAuthors acknowledge funding from the Council of Scientific and Industrial Research (CSIR), India through Grant No. BSC0212 (Wellness Genomics Project) granted to SS and VS.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nAuthors acknowledge help from Dr. Vamsi Krishna for preparation of the manuscript.\n\n\nReferences\n\nChung HJ, Uitto J: Type VII collagen: the anchoring fibril protein at fault in dystrophic epidermolysis bullosa. Dermatol Clin. 2010; 28(1): 93–105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSawyer SL, Hartley T, Dyment DA, et al.: Utility of whole-exome sequencing for those near the end of the diagnostic odyssey: time to address gaps in care. Clin Genet. 2016; 89(3): 275–84. PubMed Abstract | Publisher Full Text\n\nGupta A, Sharma YK, Vellarikkal SK, et al.: Whole-exome sequencing solves diagnostic dilemma in a rare case of sporadic acrokeratosis verruciformis. J Eur Acad Dermatol Venereol. 2016; 30(4): 695–7. PubMed Abstract | Publisher Full Text\n\nVellarikkal SK, Patowary A, Singh M, et al.: Exome sequencing reveals a novel mutation, p.L325H, in the KRT5 gene associated with autosomal dominant Epidermolysis Bullosa Simplex Koebner type in a large family from western India. Hum Genome Var. 2014; 1: 14007. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiller SA, Dykes DD, Polesky HF: A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 1988; 16(3): 1215. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H, Durbin R: Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009; 25(14): 1754–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRimmer A, Phan H, Mathieson I, et al.: Integrating mapping-, assembly- and haplotype-based approaches for calling variants in clinical sequencing applications. Nat Genet. 2014; 46(8): 912–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu J, Ng PC: SIFT Indel: predictions for the functional effects of amino acid insertions/deletions in proteins. PLoS One. 2013; 8(10): e77940. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNg PC, Henikoff S: SIFT: Predicting amino acid changes that affect protein function. Nucleic Acids Res. 2003; 31(13): 3812–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiremagalore R, Kubba A, Bansel S, et al.: Immunofluorescence mapping in inherited epidermolysis bullosa: a study of 86 cases from India. Br J Dermatol. 2015; 172(2): 384–91. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14175",
"date": "06 Jun 2016",
"name": "Robert Sidbury",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThough the authors distinguish dominant (DDEB) and recessive (RDEB) subtypes of dystrophic epidermolysis bullosa (DEB) at the outset in their abstract, they lapse into speaking of DEB in more monolithic terms later and conflating findings that are specific in some cases only to one subtype. There is a reason for this as there is considerable overlap but there are differences particularly with certain features like the propensity for developing skin cancer. The authors state \"the development of skin cancer....in almost all affected members in the third or fourth decade of life.\" This is true for RDEB phenotype but not DDEB in whom the development of squamous cell carcinoma as well as the distinctive psuedosyndactyly type of scarring much less common. This may not be the forum for parsing such details but this struck me.\nSimilarly, while I realize this is not an EB review article a brief internal reference to the Vander Oever article might allow interested readers easy access to a therapeutic update.\nFinally, I would want someone other than I with expertise in the genetic methods used to weigh in on their suitability. The methods and results to my untrained --relative to a geneticist--- eye appear sound.\nOtherwise I approve indexing this article.",
"responses": [
{
"c_id": "2044",
"date": "05 Jul 2016",
"name": "Vinod Scaria",
"role": "Author Response",
"response": "Dear Robert Sidbury, Thank you for reviewing our article and your valuable comments. We have incorporated your suggestions in the recent version. Thanks and Regards Authors"
}
]
},
{
"id": "14364",
"date": "14 Jun 2016",
"name": "Regina Fölster-Holst",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe a 5 year old boy with dystrophic epidermolysis bullosa (DEB), which is due to a novel mutation in the COL7A1 gene. The case is the first report of a genetically characterized case of DEB from India.\nIt is a well written manuscript which is worthy to be indexed with F1000Research. To understand the new mutation as the cause of severe DEB in the boy, it would be interesting to know what are the differences to other mutations of the COL7A1 gene. In other words, what does this mean on the protein level?",
"responses": []
},
{
"id": "14177",
"date": "27 Jun 2016",
"name": "Mohamed Badawy Hassan Tawfik Abdel-Naser",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nApart from few typo errors, the manuscript is well written and informative.\n\nThere are few too strong statements, e.g., \"The disease also predisposes individuals to development of skin cancer and it is estimated that almost all affected members develop cancers in the third or fourth decade of life\". Perhaps authors can insert a reference (s) that supports this statement.\n\nFig. 1. needs some corrections. Please insert (a) to refer to the upper limb, thorax and abdomen (not hands and thoracic region). b. should refer to legs and feet.\n\nIs the mentioned \"our internal cohort of 122 exomes\" published or available online? Furthermore, they may mention some of the weaknesses of the Nextera platform, such as coverage bias.\n\nIn the discussion sections, authors may mention the clinical relevance of their finding.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-900
|
https://f1000research.com/articles/5-1574/v1
|
05 Jul 16
|
{
"type": "Method Article",
"title": "An open RNA-Seq data analysis pipeline tutorial with an example of reprocessing data from a recent Zika virus study",
"authors": [
"Zichen Wang",
"Avi Ma'ayan",
"Zichen Wang"
],
"abstract": "RNA-seq analysis is becoming a standard method for global gene expression profiling. However, open and standard pipelines to perform RNA-seq analysis by non-experts remain challenging due to the large size of the raw data files and the hardware requirements for running the alignment step. Here we introduce a reproducible open source RNA-seq pipeline delivered as an IPython notebook and a Docker image. The pipeline uses state-of-the-art tools and can run on various platforms with minimal configuration overhead. The pipeline enables the extraction of knowledge from typical RNA-seq studies by generating interactive principal component analysis (PCA) and hierarchical clustering (HC) plots, performing enrichment analyses against over 90 gene set libraries, and obtaining lists of small molecules that are predicted to either mimic or reverse the observed changes in mRNA expression. We apply the pipeline to a recently published RNA-seq dataset collected from human neuronal progenitors infected with the Zika virus (ZIKV). In addition to confirming the presence of cell cycle genes among the genes that are downregulated by ZIKV, our analysis uncovers significant overlap with upregulated genes that when knocked out in mice induce defects in brain morphology. This result potentially points to the molecular processes associated with the microcephaly phenotype observed in newborns from pregnant mothers infected with the virus. In addition, our analysis predicts small molecules that can either mimic or reverse the expression changes induced by ZIKV. The IPython notebook and Docker image are freely available at: http://nbviewer.jupyter.org/github/maayanlab/Zika-RNAseq-Pipeline/blob/master/Zika.ipynb and https://hub.docker.com/r/maayanlab/zika/.",
"keywords": [
"Systems biology",
"bioinformatics pipeline",
"gene expression analysis",
"RNA-seq"
],
"content": "Introduction\n\nThe increase in awareness about the irreproducibility of scientific research requires the development of methods that make experimental and computational protocols easily repeatable and transparent1. The advent of interactive notebooks for data analysis pipelines significantly enhances the recording and sharing of data, source code, and figures2. In a subset of recent publications, an interactive notebook was published alongside customary manuscripts3. Similarly, here we present an interactive IPython notebook (http://nbviewer.jupyter.org/github/maayanlab/Zika-RNAseq-Pipeline/blob/master/Zika.ipynb) that serves as a tutorial for performing a standard RNA-seq pipeline. The IPython notebook pipeline provides scripts (http://dx.doi.org/10.5281/zenodo.56311) that process the raw data into interactive figures and permits other downstream analyses that can enable others to quickly and properly repeat our analysis as well as extract knowledge from their own data. As an example, we applied the pipeline to RNA-seq data from a recent publication where human induced pluripotent stem cells were differentiated to neuronal progenitors and then infected with Zika virus (ZIKV)4. The aim of the study was to begin to understand the molecular mechanisms that induce the observed devastating phenotype of newborn-microcephaly from pregnant mothers infected with the virus.\n\n\nMethods and results\n\nThe first publicly available study profiling gene expression changes after ZIKV infection of human cells was deposited into NCBI's Gene Expression Omnibus (GEO) in March 2016. The raw data is available (ftp://ftp-trace.ncbi.nlm.nih.gov/sra/sra-instant/reads/ByStudy/sra/SRP/SRP070/SRP070895/) from the Sequence Read Archive (SRA) with accession number GSE78711. In this study, gene expression was measured by RNA-seq using two platforms: MiSeq and NextSeq4 in duplicates. The total number of samples is eight, with four untreated samples and four infected samples. We first downloaded the raw sequencing files from SRA and then converted them to FASTQ files. Quality Control (QC) for the RNA-Seq reads was assessed using FastQC5. The reports generated by FastQC were in HTML format and can be accessed through hyperlinks from the IPython notebook. The reads in the FASTQ files were aligned to the human genome with Spliced Transcripts Alignment to a Reference (STAR)6. STAR is a leading aligner that accomplishes the alignment step faster and more accurately than other current alternatives6. We next applied featureCounts7 to assign reads to genes, and then applied the edgeR Bioconductor package8 to compute counts per million (CPM) and reads per kilobase million (RPKM). The next steps are performed in Python within the IPython notebook. We first filtered out genes that are not expressed or lowly expressed. Subsequently, we performed principal component analysis (PCA) (Figure 1). The PCA plots show that the samples cluster by infected vs. control cells, but also by platform. Next, we visualized the 800 genes with the largest variance using an interactive hierarchical clustering (HC) plot (Figure 2). This analysis separates the groups of genes that are differentially expressed by infected vs. control from those that are differential by platform. The visualization of the clusters is implemented with an interactive external web-based data visualization tool called clustergrammer (http://amp.pharm.mssm.edu/clustergrammer/). Clustergrammer provides interactive searching, sorting and zooming capabilities.\n\nZIKV-infected and mock-treated cells are colored in orange and blue, respectively. The shapes of the dots indicate the sequencing platforms: MiSeq – squares, and NextSeq - circles.\n\nThe CPM of 800 genes with the largest variance across the eight samples were log transformed and z-score normalized across samples. Blue indicates low expression and red high.\n\nThe following step is to identify the differentially expressed genes (DEG) between the two conditions. This is achieved with a unique method we developed called the Characteristic Direction (CD)9. The CD method is a multivariate method that we have previously demonstrated to outperform other leading methods that compute differential expression between two conditions9. Once we have ranked the lists of DEG, we submit these for signature analysis using two tools: Enrichr10 and L1000CDS211. Enrichr queries the up and down gene sets against over 180,000 annotated gene sets belonging to 90 gene set libraries covering pathway databases, ontologies, disease databases, and more10. The results from this enrichment analysis confirm that the downregulated genes after ZIKV infection are enriched for genes involved in cell cycle-related processes (Figure 3a). These genes are enriched for targets of the transcription factors E2F4 and FOXM1 (Figure 3b). Both transcription factors are known to regulate cell proliferation and play central role in many cancers. The downregulation of cell cycle genes was already reported in the original publication; nevertheless, we obtained more interesting results for the enriched terms that appeared most significant for the upregulated genes. Particularly, the top two terms from the mouse genome informatics (MGI) Mammalian Phenotype Level 4 library are abnormal nervous system (MP0003861) and abnormal brain morphology (MP0002152) (Table S1). This library associates gene knockouts in mice with mammalian phenotypes. These enriched terms enlist a short set of genes that potentially link ZIKV infection with the concerning observed microcephaly phenotype. Finally, to identify small molecules that can potentially either reverse or mimic ZIKV-induced gene expression changes, we query the ZIKV-induced signatures against the LINCS L1000 data. For this, we utilize L1000CDS211, a search engine that prioritize small molecules given a gene expression signature as input. L1000CDS2 contains 30,000 significant signatures that were processed from the LINCS L1000 data with the CD method. The results suggest small molecules that could be tested in follow-up studies in human cells for potential efficacy against ZIKV (Table S2).\n\nBar plots of the top enriched gene sets from the (a) ChEA and (b) KEGG libraries for the downregulated genes after ZIKV infection.\n\nTo ensure the reproducibility of the computational environment used for the whole RNA-Seq pipeline, we packaged all the software components used in this tutorial, including the command line tools, R packages, and Python packages into a Docker image. This Docker image is made publically available at https://hub.docker.com/r/maayanlab/zika/. The Docker image was created based on the specifications outlined on the official IPython’s Scipy Stack image (https://hub.docker.com/r/ipython/scipystack/). The additional command line tools, R scripts, and Python packages together with their dependencies were compiled and installed into the Docker image. The RNA-Seq pipeline Docker image was deployed onto our Mesos cluster, which allows users to run the IPython notebook interactively. The Docker image can also be downloaded and executed on local computers and servers, or deployed in the cloud if users have access to cloud provider services with a Docker Toolbox installed (https://www.docker.com/products/docker-toolbox). We also provide detailed instructions on how to download and execute the Docker image (https://hub.docker.com/r/maayanlab/zika/).\n\nThe ‘Dockerization’ of the RNA-Seq pipeline facilitates reproducibility of the pipeline at the software level because the Docker image ensures that all versions of the software components are consistent and static. Dockerization also helps users to handle the complex installation of many dependencies required for the computational pipeline. Moreover, the Docker image can be executed on a single computer, clusters/servers and on the cloud. The only limitation of having a Docker image is that it prevents users from adding or altering the various steps which require additional software components and packages. However, advanced users can build their own Docker images based on our initial image to customize it for their needs.\n\n\nDiscussion and conclusions\n\nIn summary, we provide an open source RNA-seq processing pipeline (Figure 4) that can be used to extract knowledge from any study that profiled gene expression using RNA-seq applied to mammalian cells, comparing two conditions. The advantage of providing the pipeline in the IPython notebook format and as a Docker container is that it enables others to quickly reproduce our results with minimal overhead and potentially apply similar methodology for the analysis of other similar datasets. Advanced users can add, improve and customize the pipeline by forking it on GitHub. The results that we obtained for ZIKV are consistent with the results published in the original study, but also enhance those findings by discovering a link between the upregulated genes and genes that, when knocked out in mice, induce morphological brain defects. Some of these genes could be the causal genes of the microcephaly phenotype observed in newborns of mothers infected with the virus. Nevertheless, caution should be used when interpreting these results because they may simply indicate a reduction in cell cycle activity and an increase in neuronal differentiation of the type of cells used in the original study.\n\n\nData and software availability\n\nThe IPython notebook, as well as other scripts and data files for this tutorial are available on GitHub at: https://github.com/MaayanLab/Zika-RNAseq-Pipeline, doi: http://dx.doi.org/10.5281/zenodo.5631112.\n\nThe Docker image for this tutorial is available on DockerHub at: https://hub.docker.com/r/maayanlab/zika/.",
"appendix": "Author contributions\n\n\n\nAM conceived and lead the study. ZW developed the software and performed the analysis. AM interpreted the results. Both authors wrote the paper and tutorials.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work is partially supported by the National Institutes of Health (NIH) grants U54HL127624, U54CA189201, and R01GM098316.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Dr. Ajay Pillai from NHGRI for useful suggestions and Kathleen Jagodnik from NASA for copyediting an early version of the manuscript.\n\n\nSupplementary material\n\nTable S1.\n\nTop enriched gene sets from the MGI Mammalian Phenotype Level 4 gene set library for the upregulated genes after ZIKV infection.\n\nClick here to access the data.\n\nTable S2.\n\n(a) Top scoring small molecules that are potential mimickers of the ZIKV infection signatures. (b) Top scoring small molecules that are potential reversers of the ZIKV infection signatures.\n\nClick here to access the data.\n\n\nReferences\n\nBegley CG, Ellis LM: Drug development: Raise standards for preclinical cancer research. Nature. 2012; 483(7391): 531–533. PubMed Abstract | Publisher Full Text\n\nShen H: Interactive notebooks: Sharing the code. Nature. 2014; 515(7525): 151–152. PubMed Abstract | Publisher Full Text\n\nhttps://IPython.org. A gallery of interesting IPython Notebooks. [cited 2016 25 April]; 2016. Reference Source\n\nTang H, Hammack C, Ogden SC, et al.: Zika Virus Infects Human Cortical Neural Progenitors and Attenuates Their Growth. Cell Stem Cell. 2016; 18(5): 587–590. PubMed Abstract | Publisher Full Text\n\nAndrews S: FastQC: A quality control tool for high throughput sequence data. 2010. Reference Source\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014; 30(7): 923–930. PubMed Abstract | Publisher Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClark NR, Hu KS, Feldmann AS, et al.: The characteristic direction: a geometrical approach to identify differentially expressed genes. BMC Bioinformatics. 2014; 15(1): 79. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen EY, Tan CM, Kou Y, et al.: Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics. 2013; 14(1): 128. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuan Q, Reid SP, Clark NR, et al.: L1000CDS 2: LINCS L1000 Characteristic Direction Signature Search Engine. NPJ Systems Biology and Applications, 2016. Reference Source\n\nWang Z, Ma'ayan A: Zika-RNAseq-Pipeline v0.1. Zenodo. 2016. Data Source"
}
|
[
{
"id": "14924",
"date": "21 Jul 2016",
"name": "Ravi K. Madduri",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWang and Ma’ayan introduced an RNA-seq pipeline tutorial using IPython notebook and a Docker image. Specifically, the authors applied the pipeline to analyze data from a recent Zika virus study. The authors found that their pipeline not only confirms the down-regulated cell cycle genes, but also uncovers a set of genes with a biological function that potentially associated with a particular phenotype. While the work and context sound interesting, there are several concerns that need to be addressed or discussed:\n\nThis reviewer really liked the approach that the authors have taken to showcase analysis. I wish more researchers adopt this methodology.\n\nThis could be a great way to do additional analysis easily.. I wonder if authors can look into additional RNASeq pipelines and compare/contrast how Jupiter-friendly they are.\n\nMinor comments:\nThe clustering (in Figure 1) is based on z-score and the 800 genes serve well to cluster the samples into two different groups. Was z-scores close to zero excluded as they are uninformative? Was FDR applied?",
"responses": []
},
{
"id": "15290",
"date": "29 Jul 2016",
"name": "Apostolos Zaravinos",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors provide an open source RNA-seq processing pipeline that can be used to extract differential expression data between two conditions from an RNA-seq experiment. To test this, data from a recent publication where human iPSCs were differentiated to neuronal progenitors and then infected with Zika virus (ZIKV), were analysed with their pipeline, and their results were consistent with the original ones. Of interest, the authors discovered a link between the upregulated genes of this study and genes that, when knocked out could cause microcephaly observed in newborns of mothers infected with ZIKV. Overall, this is a novel and exciting computational protocol that promotes reproducibility and transparency of the results, and it is definately worth to be tested in other conditions, as well (eg, cancer).\nMinor comment: Perhaps its a matter of zoom-in/out of the clustergrammer, but it would be nice to show the names of the top up-/down-regulated genes (or their categories) as depicted in Figure 2.",
"responses": []
},
{
"id": "15289",
"date": "15 Aug 2016",
"name": "Fredrik Pettersson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWang and Ma’ayan have developed an open source RNAseq processing pipeline using standard methods as well as integrated tools for visualization and data analysis of differentially expressed genes between two sets of data. They test the pipeline by reanalyzing a previously published study of Zika virus infected human cells. They replicated the main result of the original study where genes related to the cell cycle were downregulated after infection. They also find a potential link to genes involved in brain morphology and a normal functioning of the nervous system in mice, something the original study missed. These genes are upregulated after Zika virus infection. This pipeline should be useful in any type of study where two conditions are compared, e.g. infected vs uninfected cells or treated vs untreated condition.\nMinor comments:\nFigure 2. Have the different conditions/figure labels been mixed up? The 2 Zika infected MiSeq conditions look complementary to the 2 mock infected NextSeq controls, while the 2 Zika infected NextSeq conditions look complementary to the 2 mock infected MiSeq controls.\n\nFigure 3a and 3b have been mixed up. 3b should be 3a and vice versa.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1574
|
https://f1000research.com/articles/4-1197/v1
|
02 Nov 15
|
{
"type": "Case Report",
"title": "Case Report: Microsurgical excision of grade 5 cerebral AVM",
"authors": [
"Sunil Munakomi",
"Binod Bhattarai",
"Iype Cherian",
"Binod Bhattarai",
"Iype Cherian"
],
"abstract": "In this case report, we discuss the microsurgical management of a Spetzler-Martin grade 5 arteriovenous malformation (AVM) in a young boy who presented with a hemorrhagic episode and had a high calculated risk of rebleeding. We also outline the rationale for choosing the management option.",
"keywords": [
"arteriovenous malformation",
"AVM",
"grading",
"management"
],
"content": "Introduction\n\nAs of now, many tenets exist regarding management of high grade cerebral arterio-venous malformation (AVM) management, making a rigid algorithm impossible to create. In experienced hands, microsurgery proved to have better results, compared to other treatments1,2. Herein, we report a microsurgical management of a grade 5 arteriovenous malformation (AVM) in a young patient with a high predicted risk for rebleeding.\n\n\nCase report\n\nA 22-year-old Brahmin male from Khaireni, a remote village in Nepal, presented to our emergency room with a sudden-onset severe headaches and left-sided weakness over the last 24 hours. Physical examination revealed a Glasgow coma scale (GCS) of 14/15 with left-sided hemiparesis of 3+/5(Medical research council grading). Medical history was significant for a few episodes of paroxysmal headaches since last couple of years, which improved after taking 500 mg Paracetamol tablet on an ‘as needed’ basis. The frequency and intensity of the headache had worsened in the last few months. There was no significant family history. An urgent head computerized tomogram (CT) revealed evidence of a hyperdense lesion with peripheral stippled calcification on the right parietal region (Figure 1). There was also coating of vessel along the middle cerebral artery (MCA) territory (Figure 2) and hyperdensity along the deep venous territory. A four-vessel diagnostic carotid angiography revealed Grade 5 Spetzler-Martin AVM in the right insular territory with feeders from lenticulostiates of the middle cerebral artery (Figure 3). Drainage was to the deep draining veins and also to the superior sagittal sinus (Figure 4).\n\nMultiple factors such as young age at presentation, the fact that the lesion had bled, presentation of patient with deficits associated with the lesion on the non-dominant side, presence of deep venous drainage and intra-nidal aneurysm led to a high calculated risk for rebleeding in the patient. We therefore decided on surgical management, despite the high grade of the lesion. After explanation of the risks of the treatment and role of adjuvants in the form of radiosurgery and embolisation the patient was taken up for microsurgical excision. Since the facility of radiosurgery is not available in the country, we only had the option of embolisation of the feeders prior to the surgical excision of the lesion. However, since the lesion had only low velocity feeders from the lenticulostriate vessels, we opted for direct microsurgical management. After a liberal craniotomy, basal cisterns were opened to gain access to the M1 branch of the MCA. We identified the major deep draining vein that was looping over the MCA bifurcation with the help of Indocyanine Green (ICG) venography. We placed a temporary clip on M1, then made a minimal corticostomy over the parietal cortex and continued our dissection over the gliotic tissue surrounding the AVM taking care of the minimal bleeders with the help of bipolar cauterization and avoiding inadvertent entry to the nidus. Lastly a clip was applied to the draining vein after completely dissecting the AVM nidus. The lesion was finally excised (Figure 5). Complete hemostasis was confirmed.\n\nPostoperatively his blood pressure was rigorously monitored so as not to overshoot the mean arterial pressure above 100 mm of mercury so as to prevent breakthrough perfusion rebleeding. Patient was started on Sodium Valproate (1 gm stat followed by 300 mg IV 8 hourly) and Nimodipine (60 mg 4 hourly via nasogastric tube) for seizure and vasospasm prophylaxis, respectively. Repeat head CT scan the following morning revealed no cavity hematoma or any evidence of vasospasm (Figure 6). Patient was extubated uneventfully. He had hemiparesis of 3+ in upper limbs and 3 in lower limbs. Patient was started on physiotherapy and finally discharged home on the 7TH post-operative day after removal of sutures. Patient came for follow-up 2 weeks later walking on his own with left upper limb weakness of grade 3+/5. The Nimodipine was tapered off in the subsequent three weeks. The patient was advised to continue Na Valproate 300 mg orally three times a day for at least a year.\n\n\nDiscussion\n\nBleeding within the AVM is considered a significant predictor of rebleeding. Other important factors moderating risk of rebleeding include deep venous drainage3. Studies have verified that the risk of rebleeding under these circumstances is as high as 34.4% compared to just 0.9% per year in patients without these risk factors4,5. Another important factor to be considered while calculating the risk of rebleeding is the presence of concurrent aneurysm within the AVM (6.93% with aneurysm Vs 3.99% without aneurysm)3.\n\nUp to 40% of cases with AVM manifest neurological deficits6, mostly attributable to hemorrhage. A minority of only 5% to 15% of such deficits are related to factors such as coronary steal phenomenon and venous hypertension7–9.\n\nThe Spetzler-Martin Scale is used to estimate the risk of surgical resection of an AVM with higher grades being associated with greater surgical morbidity and mortality10. Multivariate studies have shown this grading system to reliably predict permanent major morbidity or mortality at the following levels: Grade I (4%), Grade II (10%), Grade III (18%), Grade IV (31%), and Grade V (37%)11. This data has been further validated prospectively, and this grading system remains the most widely used among neurosurgeons and neurointerventionalists12.\n\nHan et al. reported the management of 73 grade 4 and 5 lesions and found the annual hemorrhage rate for untreated lesions to be only 1.5% versus 10.4% for partially treated lesions13. Grade IV or V lesions are only treated in circumstances of progressive neurological deterioration from hemorrhage, vascular steal, or seizure as seen in our case, which had a high risk of rebleeding because of presentation at young age with hemorrhagic episode, large size of the nidus, deep venous drainage pattern and associated aneurysm within the AVM.\n\nThere is time-lag of about two years following radiosurgery for complete obliteration of the nidus in the lesion. The risk of hemorrhage in this time period is around 4.8% per year14 which parallels the natural history of the lesion after bleeding. However there is a risk of inadvertent radiation injury to the adjacent eloquent brain area15 and also a risk of symptomatic radiation necrosis in around 9% of cases15,16.\n\nThe main indication for other embolisation options in such a high grade of AVM is in order to downgrade the lesion and to minimise the intraoperative blood loss so as to make the lesion amenable for microsurgical excision, which bears an acceptable complication rate of around just 6.5%17. One study has shown that the deep venous drainage, higher grade of the lesions and the periprocedural hemorrhage are predictors of post procedural complications following the embolisation treatment17.\n\nIn our case there were only few feeders from the lenticulostriate branches from MCA: not ideal for embolization. Partial embolisation of the lesion will not reduce the risk of hemorrhage to zero3. Partial embolisation of the high grade lesions are only justified in few circumstances, such as in vascular steal phenomenon or an AVM with associated aneurysm18.\n\n\nConclusion\n\nIn a few selected cases who have a high calculated risk of rebleeding, microsurgical excision remains a therapeutic option even for a high grade AVM especially in centers with limited resources for intervention and radiosurgery. However, all the patients should be well counseled about the available alternative mode of intervention and the associated risks. The management plan in each patient should be tailored addressing factors such as age of the patient, mode of presentation, grade of the lesion, treatment modalities and expertise availability etc.\n\n\nConsent\n\nWritten, informed consent was sought and attained from the father of the patient as per medical protocol in Nepal.",
"appendix": "Author contributions\n\n\n\nDr Sunil prepared the manuscript and obtained the pictures. Dr Binod and Dr Cherian revised and confirmed the final manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declared no conflict of interest.\n\n\nGrant information\n\nThe authors declared that no funding was involved in supporting this work.\n\n\nReferences\n\nPotts MB, Chang EF, Young WL, et al.: Transsylvian-transinsular approaches to the insula and basal ganglia: operative techniques and results with vascular lesions. Neurosurgery. 2012; 70: 824–834; discussion 834. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPotts MB, Young WL, Lawton MT, et al.: Deep arteriovenous malformations in the Basal Ganglia, thalamus, and insula: microsurgical management, techniques, and results. Neurosurgery. 2013; 73: 417–427. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Costa L, Wallace MC, Ter Brugge KG, et al.: The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke. 2009; 40(1): 100–105. PubMed Abstract | Publisher Full Text\n\nHofmeister C, Stapf C, Hartmann A, et al.: Demographic, morphological, and clinical characteristics of 1289 patients with brain arteriovenous malformation. Stroke. 2000; 31(6): 1307–1310. PubMed Abstract | Publisher Full Text\n\nPollock BE, Flickinger JC, Lunsford LD, et al.: Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke. 1996; 27(1): 1–6. PubMed Abstract | Publisher Full Text\n\nThe Arteriovenous Malformation Study Group: Arteriovenous malformations of the brain in adults. N Engl J Med. 1999; 340(23): 1812–1818. PubMed Abstract | Publisher Full Text\n\nMast H, Mohr JP, Osipov A, et al.: ‘Steal’ is an unestablished mechanism for the clinical presentation of cerebral arteriovenous malformations. Stroke. 1995; 26(7): 1215–1220. PubMed Abstract | Publisher Full Text\n\nChoi JH, Mast H, Sciacca RR, et al.: Clinical outcome after first and recurrent hemorrhage in patients with untreated brain arteriovenous malformation. Stroke. 2006; 37(5): 1243–1247. PubMed Abstract | Publisher Full Text\n\nFriedlander RM: Clinical practice. Arteriovenous malformations of the brain. N Engl J Med. 2007; 356(26): 2704–2712. PubMed Abstract | Publisher Full Text\n\nSpetzler RF, Ponce FA: A 3-tier classification of cerebral arteriovenous malformations. Clinical article. J Neurosurg. 2011; 114(3): 842–9. PubMed Abstract | Publisher Full Text\n\nHamilton MG, Spetzler RF: The prospective application of a grading system for arteriovenous malformations. Neurosurgery. 1994; 34(1): 2–6; discussion 6–7. PubMed Abstract | Publisher Full Text\n\nCastel JP, Kantor G: Postoperative morbidity and mortality after microsurgical exclusion of cerebral arteriovenous malformations. Current data and analysis of recent literature. Neurochirurgie. 2001; 47(2–3 Pt 2): 369–383. PubMed Abstract\n\nHan PP, Ponce FA, Spetzler RF: Intention-to-treat analysis of Spetzler-Martin grades IV and V arteriovenous malformations: natural history and treatment paradigm. J Neurosurg. 2003; 98(1): 3–7. PubMed Abstract | Publisher Full Text\n\nHernesniemi JA, Dashti R, Juvela S, et al.: Natural history of brain arteriovenous malformations: a long-term follow-up study of risk of hemorrhage in 238 patients. Neurosurgery. 2008; 63(5): 823–829. PubMed Abstract | Publisher Full Text\n\nFlickinger JC, Lunsford LD, Kondziolka D, et al.: Radiosurgery and brain tolerance: an analysis of neurodiagnostic imaging changes after gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys. 1992; 23(1): 19–26. PubMed Abstract | Publisher Full Text\n\nFinitsis S, Anxionnat R, Bracard S, et al.: Symptomatic Radionecrosis after AVM Stereotactic Radiosurgery. Study of 16 Consecutive Patients. Interv Neuroradiol. 2005; 11(1): 25–33. PubMed Abstract | Free Full Text\n\nLedezma CJ, Hoh BL, Carter BS, et al.: Complications of cerebral arteriovenous malformation embolization: multivariate analysis of predictive factors. Neurosurgery. 2006; 58(4): 602–611; discussion 602–11. PubMed Abstract | Publisher Full Text\n\nKrings T, Hans FJ, Geibprasert S, et al.: Partial “targeted” embolisation of brain arteriovenous malformations. Eur Radiol. 2010; 20(11): 2723–31. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "10999",
"date": "12 Nov 2015",
"name": "Kenichiro Kikuta",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors reported a case of S-M grade 5 AVM successfully treated by microsurgical resection. I have some questions about as below.How did the authors measure the size of AVM. How did they distinguish hematoma from AVM? Was the AVM alone really above 6 cm in diameter? The authors should show the image such as T2-weighted images of MRI or enhanced CT clearly indicating the size of AVM. Was the AVM really located in the eloquent area? The authors should show the location of nidus clearly. I think not nidus but only hematoma was located in the insult. Magnification of figures 3 and figures 4 was too high. The authors should show angiograms of A-P view and lateral view with less magnification. Why didn’t the authors perform preoperative embolization? I think it would be helpful in cases of high-grade AVMs.",
"responses": []
},
{
"id": "12755",
"date": "22 Mar 2016",
"name": "Sunil Kumar Singh",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is a well written case report and the discussion is precise. Main points for not approving this article are:Grading appears faulty as size of nidus, eloquence and drainage are suspicious on single images. more images might have been helpful. The AVM at best appears to be grade 3. As such, it would be a routine microsurgical excision of an AVM",
"responses": []
}
] | 1
|
https://f1000research.com/articles/4-1197
|
https://f1000research.com/articles/5-1565/v1
|
01 Jul 16
|
{
"type": "Research Article",
"title": "Centromere detection of human metaphase chromosome images using a candidate based method",
"authors": [
"Akila Subasinghe",
"Jagath Samarabandu",
"YanXin Li",
"Ruth Wilkins",
"Farrah Flegal",
"Joan H. M. Knoll",
"Peter K. Rogan",
"Akila Subasinghe",
"YanXin Li",
"Ruth Wilkins",
"Farrah Flegal",
"Joan H. M. Knoll"
],
"abstract": "Accurate detection of the human metaphase chromosome centromere is a critical element of cytogenetic diagnostic techniques, including chromosome enumeration, karyotyping and radiation biodosimetry. Existing centromere detection methods tends to perform poorly in the presence of irregular boundaries, shape variations and premature sister chromatid separation. We present a centromere detection algorithm that uses a novel contour partitioning technique to generate centromere candidates followed by a machine learning approach to select the best candidate that enhances the detection accuracy. The contour partitioning technique evaluates various combinations of salient points along the chromosome boundary using a novel feature set and is able to identify telomere regions as well as detect and correct for sister chromatid separation. This partitioning is used to generate a set of centromere candidates which are then evaluated based on a second set of proposed features. The proposed algorithm outperforms previously published algorithms and is shown to do so with a larger set of chromosome images. A highlight of the proposed algorithm is the ability to rank this set of centromere candidates and create a centromere confidence metric which may be used in post-detection analysis. When tested with a larger metaphase chromosome database consisting of 1400 chromosomes collected from 40 metaphase cell images, the proposed algorithm was able to accurately localize 1220 centromere locations yielding a detection accuracy of 87%.",
"keywords": [
"methaphase chromosome",
"centromere detection",
"algorithm"
],
"content": "Introduction\n\nThe centromere of a human chromosome (Figure 1) is the primary constriction to which the spindle fiber is attached during the cell division cycle (mitosis). The detection of this salient point is the key to calculating the centromere index which can lead to the type and the number of a given chromosome. The reliable detection of the centromere by image analysis techniques is challenging due to the high morphological variations of chromosomes on microscope slides. This variation is caused by various cell preparation and staining methods along with other factors that occur during mitosis. Irregular boundaries and large variations in chromosome morphology can cause a detection algorithm to miss the constriction, especially in high resolution chromosomes. Premature sister chromatid separation can also pose a significant challenge, since the degree of separation can vary from cell to cell, and even among chromosomes in the same cell. In such cases, the width constriction can be missed by image processing algorithms, and can result in incorrect localization of a centromere on one of the sister chromatids.\n\nFrom an image analysis perspective, the high morphological variations in human chromosomes, due to their non rigid nature, pose a significant challenge. Cell preparation and staining techniques also vary among the laboratories. The end results obtained from clinical cytogenetic vs. reference biodosimetry laboratories can produce chromosome images that differ significantly in their appearance. As an example, chromosomes that were DAPI (4’,6-diamidino-2-phenylindole) stained shows different intensity features and boundary characteristics from chromosomes subjected only to Giemsa staining. Additionally, the stage of metaphase in which the cells were arrested along with environmental factors such as humidity during slide preparation can dictate the shape characteristics of individual cells and introduce a large variance to the data set. Furthermore, in some preparation methods, the cells are denatured, causing the detected chromosome boundary to be erratic. These same factors can also dictate the amount of premature sister chromatid separation in some of the cells. Effective algorithms for centromere detection need to be able to handle the high degree of shape variability present in different chromosomes, while correcting for artifacts such as premature sister chromatid separation. Figure 2 illustrates a sample set of shapes of chromosomes in the data set and their high morphological variations.\n\nDepicts various degrees of sister chromatid separation present in some Giemsa stained chromosome images (Figure 2(a)–(c)) as well as some longer chromosomes characteristics of those prepared at a clinical cytogenetic laboratory (Figure 2(d)–(f)).\n\nThis research forms an essential component of detecting dicentric chromosomes (possessing two centromeres) which is used as a diagnostic test of radiation exposures in cytogenetic biodosimetry. The ability of the proposed algorithm to handle high degrees of morphological variation and also to detect and correct for the artifact created by premature sister chromatid separation in cell images is also critical to detecting dicentric chromosomal abnormalities.\n\nNumerous computer algorithms have been proposed over time for chromosome analysis ranging from metaphase finding1, karyotype analysis2 to centromere and dicentric detection3,4. These methods are either constrained by the protocol used for staining the cell image or by the morphology of the chromosome. We have previously proposed an algorithm to locate the centromere by calculating a centerline with no spurious branches irrespective of boundary irregularities or the morphology of the chromosome5. This was later improved by using a Laplacian-based width-profile generation algorithm that integrates intensity measurements in a weighting scheme, biases the thickness measurement by tracing vectors across regions of homogeneous intensity6. Mohammad proposed an approach where he used our previous approach to derive the centerline and then used a curvature measure to localize the centromere location instead of the width measurements7. Another interesting approach by Jahani and Setarehdan involves artificially straightening chromosomes prior to creating the trellis structure using the centerline derived through morphological thinning8. Yet all these methods, including our previous approach, work well only with smooth object boundaries. The absence of a smooth boundary will directly affect the centerline and thus make the feature calculations noisy. Furthermore, the accuracy of all these methods is adversely impacted by sister chromatid separation. Although a commercial system exists for detecting dicentric chromosomes, it is semi-automatic and requires manual review of cells9. Furthermore, no published accuracy figures for detection of centromeres exist for this system. We propose a candidate based centromere localization algorithm capable of processing highly bent chromosomes prepared with a variety of staining techniques. This method can also detect and correct for artifacts introduced by premature sister chromatid separation.\n\nTo address image processing artifacts arising from sister chromatid separation, the proposed algorithm utilizes a new contour partitioning technique which identifies the telomere regions. This partitioning technique evaluates various combinations of salient points along the chromosome boundary by using machine learning together with a specially designed set of features. The partitioned contour is then used to generate a set of centromere candidates using local minima of the width profile. These centromere candidates are then classified using machine learning with a second set of features which incorporates contour shape as well as intensity information. This paper also introduces the Candidate-Based Centromere Confidence (CBCC) metric, which we use as an indicator of confidence of the detected location of the centromere. This metric is used in tests of the algorithm on a larger data set of chromosomes, with the aim of validating the performance of the algorithm.\n\nThe following section describes the proposed algorithm in detail. In section we show how this algorithm performed with a large data set and in section we comment on the performance and how it compares with other methods.\n\n\nMethods\n\nThis section describes the proposed candidate based centromere detection algorithm in detail. This method can be functionally divided into the following steps for clarity,\n\n1. Segmentation & centerline extraction\n\n2. Contour partitioning & correcting for sister chromatid separation\n\n3. Candidate point generation & metaphase centromere detection\n\nOf these, step 1 was performed using algorithms that were published by us previously5,6. A brief description of this is included below for improved readability.\n\nThe chromosome database was created by manually selecting individual chromosomes that are well separated. During this process, images of cells with incomplete chromosome complements and those with higher densities of overlapping or touching chromosomes were discarded using a content-based classification procedure as described by others10. We have also developed Automated Dicentric Chromosome Identifier (ADCI) software which can automatically select individual chromosomes11. However, it was not used in this study.\n\nPre-processing steps for each chromosome image include application of a median filter followed by intensity normalization. The chromosome is then tresholded using Otsu’s method and the contour of that binary object is used as the starting point for Gradient Vector Flow (GVF) active contours. The use of GVF active contour algorithm produces a contour that is smooth and that converges to boundary concavities12.\n\nIn order to calculate the width profile of the chromosome using the thickness measuring algorithm, the chromosome contour is divided longitudinally into two approximately symmetric segments. We used Discrete Curve Evolution (DCE) based skeletal pruning5 to obtain an accurate centerline. DCE is a polygon evolution algorithm which evolves through vertex deletion based on a relevance measurement13. Using DCE, the chromosome boundary is reduced to the smallest possible polygon (a triangle). The shortest branch of the resulting skeleton is pruned to yield two points which belongs to the two ends (telomeres) and are used to obtain the centerline through the chromosome. These are called anchor points and denoted by EP.\n\nThroughout this paper, we use the supercript P to refer to various point sets on the chromosome object contour C ∈ ℝ2. This set of points is used for contour partitioning in the next section.\n\nSister chromatid separation in chromosomes is an integral process that occurs during the metaphase stage of mitosis. Depending on the stage of mitosis at which the cells were arrested, varying degrees of sister chromatid separation may be evident. Furthermore long exposure to colcemid, a chemical agent which is used mainly as a preparatory chemical in biodosimetry studies to maximize the number of metaphase cells, can cause or exacerbate this condition and produce sister chromatid separation. It is important that the algorithm and associated software be able to analyze chromosomes with sister chromatid separation.\n\nAccurate partitioning of the telomere region is necessary to identify evidence of sister chromatid separation and therefore correct for any such artifact as well as to split the contour into two segments accurately. Curvature of the contour is one of the most commonly used features for detecting salient points that can be used for partitioning14. An important requirement is that the location of these salient points needs to be highly repeatable under varying levels of object boundary noise. The DCE method described in the previous section was used again to provide a set of initial salient points on the contour of the chromosome outline. This is because this method performs well with boundaries regardless of whether they are smooth or not, yielding repeatable results15. The ability to terminate the process of DCE shape evolution at a given number of vertices further lends to its applicability. It was empirically established that a termination at 6 points would ensure that the required telomere end points will be retained within the set of candidate salient points. Two of those 6 points will include the anchor points, EP obtained in the previous step (section). Contour partitioning is performed by selecting the best 4 point combination (including the two anchor points) that represents all the telomere end points.\n\nThe approach for selecting the optimal contour partitioning point combination occurs in two stages. Initially, a SVM classifier using features F1s–F11s (described below) was trained to detect and label preferred combinations from the given 12 possible combinations for each chromosome. At this stage, all the combinations across the data set are used as a pool of candidates to train the classifier. Then, the signed Euclidian distance from the separating hyperplane (say ρ) is computed for each of the candidates for a given chromosome, considering only the combinations of that chromosome. This process ranks all the candidates according to the likelihood they are a preferred candidate. Unlike traditional rule-based ranking algorithms, this approach requires very little high level knowledge of the desirable characteristics. The positioning of the separating hyperplane encapsulates this high level information through user-specified ground truth. The highest-ranked candidate is selected as the best combination of contour partitions for the given chromosome. The formal description of this procedure follows.\n\nLet Φh be the curvature value at candidate point h and S ∈ ℝ2 be the skeleton of the chromosome with 6 DCE point stop criteria. We now define the following set of points (see Figure 3),\n\nDP (⊂ C) is the set of six DCE vertices.\n\nEP is the set of two anchor points\n\nSP = DP – EP constitutes of all the points in DP except the anchor points (EP). These are the four telomere end-point candidates.\n\nThe (blue) line connects the set of points constituting the combination considered in this instance.\n\nThen the family of sets TP for all possible combinations with the sets EP and SP would contain,\n\n{E1P,S1P,E2P,S2P}, {E1P,S1P,E2P,S3P},{E1P,S1P,E2P,S4P}, {E1P,S2P,E2P,S1P},{E1P,S2P,E2P,S3P}, {E1P,S2P,E2P,S4P},{E1P,S3P,E2P,S1P}, {E1P,S3P,E2P,S2P},{E1P,S3P,E2P,S4P}, {E1P,S4P,E2P,S1P},{E1P,S4P,E2P,S2P}, {E1P,S4P,E2P,S3P}.\n\nFigure 3 illustrates one such combination where the selected (connected by the blue line segments) combination for the contour partitioning points are given by {E1P,S4P,E2P,S1P}.\n\nIn order to identify the best possible combination for contour partitioning, we have used a SVM classifier trained with the 11 different features (F1s–F11s) indicated below. Features F1s and F2s provide an indication to the saliency of the candidate point with respect to the skeletonization process. Features F3s to F5s are three normalized features which capture the positioning of each candidate in the given combination. F6s and F7s represent the shape or the morphology of the chromosome of interest (same values for all 12 combinations). The rationale behind the inclusion of these features is that they account for morphological variations across the cell images in the data set. F8s and F9s represent the curvature of the candidate points as well as the concavity/convexity of those locations. The features F10s and F11s are two Euclidean distance-based features which capture the proportion of each telomere region in the combination to the perimeter of the rectangle made by connecting the 4 candidate points. During our investigation, we observed a significant improvement of the accuracy of classification by the inclusion of these two features.\n\nLet d (p, q) denote the Euclidean distance between the points p and q. Similarly let l (p, q) represent the length of the curve between p and q, which are points from the set DP. Then, for each contour partitioning combination in TP given by {E1P,SiP,E2P,SjP} (where i and j are integer values such that 1 ≤ i, j ≤ 4 and i ≠ j), two main length measurement ratios (r1 and r2) are used for both calculating length based features, as well as for normalizing these features. r1 = l(E1P,SiP)l(E1P,SjP) yields the chromosome width/length with respect to the anchor point E1P for the given contour partitioning combination (refer Figure 3). Similarly r2 = l(E2P,SiP)l(E2P,SjP) is calculated with respect to the anchor point E2P. Then, the set of features Fs for each contour partitioning combination is defined as follows,\n\n1. F1s = 1 if the point SiP belongs to a skeletal end point (SiP ∈ (S ∩ C)). Otherwise, F1s = 0.\n\n2. F2s = 1 if the point SjP belongs to a skeletal end point (SjP ∈ (S ∩ C)). Otherwise, F2s = 0.\n\n3. F3s=[ 1−|r1−r2max(r1,r2)| ] where 0 < F3s < 1. This calculates the chromosome width/length ratio for each anchor point and the difference between the two measures. Two similar fractions would result in a high value for the feature F3s.\n\n4. F4s=[ 1−r1max(r1,r2) ] where 0 < F4s < 1. This calculates the chromosome width/length ratio with respect to the first anchor point (E1P). Except for smallest chromosomes at the highest degree of metaphase condensation, the telomere axis is shorter than the longitudinal dimension of the chromosome. Therefore, a lower length ratio measurement is a higher value for the feature F4s and is a desirable property.\n\n5. F5s=[ 1−r2max(r1,r2) ] where 0 < F5s < 1. This is same as F4s, but from the other anchor point, E2P.\n\n6. F6s : ratio of length of the chromosome to area of the chromosome. This provides a measure of elongation of a chromosome.\n\n7. F7s : ratio value of perimeter of the chromosome to the area of the chromosome. This provides a measure of how noisy the object boundaries are.\n\n8. F8s : average of the curvature values Φh of the candidates. The curvature is an important measurement of the saliency of the candidate points.\n\n9. F9s : number of the negative curvature values (Φh < 0) of the candidates points (SiP and SjP). The telomere region end points are generally characterized by points with high convexity. The number of negative angles yield how concave the points of interest are.\n\n10. F10s=d(E1P,SiP)D where D=∑x=1,y=i,jx=2d(ExP,SyP). This feature calculates the normalized Euclidean distance between the anchor point P1E and the candidate PiS which makes up one telomere region.\n\n11. F11s=d(E2P,SjP)D where D=∑x=1,y=i,jx=2d(ExP,SyP). This is the same as feature F10s, but calculated for the other anchor point.\n\nA data set of 1400 chromosomes was collected from 40 metaphase cell images, which together yield 16,800 possible combinations of feature sets for contour partitioning. Three expert cytogeneticists marked the viable combinations of the salient points that capture the telomere regions for training the SVM classifier. The procedure involved training and testing with 2 fold cross validation (50% - train data, 50% - test data). We obtained accuracy, sensitivity and specificity values of 94%, 97% and 68%, respectively. The results demonstrate the ability of the feature set to effectively detect good combinations of candidate points for partitioning telomere regions. Although the low specificity suggests that some false positive telomeres were detected, this did not affect the accuracy of the contour partitioning, since the algorithm picks the optimal combination based on its rank rather than the classification label.\n\nCorrecting the deviation of the centerline for the effects of premature sister chromatid separation can be a difficult problem to solve. Once the best combination for the end points of the telomere region is selected, the telomere portions are segmented. Premature sister chromatid separation is detected from differences in the chromosome shape in the telomere region. This problem is solved with an algorithm that creates a set of features using functional approximation of the shape characteristics unique to premature sister chromatid separation and is derived from the coefficients calculated for each telomere6. A second SVM classifier is trained on these features to effectively detect these inherent shape variations of the sister chromatids. Once identified, correction is performed by extending the sample point (on the pruned centerline) to pass through the mid point of the partitioned telomere region. By getting the contour partitioned accurately, the correction process is significantly simplified.\n\nIn a previously described candidate-based approach, four candidate points were selected based on the minima values from the width profile16. However, this limits the number of possible locations that could be detected as the centromere location. Especially in cases where a high degree of sister chromatid separation is evident, limiting the search to just few candidates can have adverse effects. Therefore, we consider all possible local minima locations as candidates for the centromere location in a given chromosome, which are selected using the simple criteria given below.\n\nOur notation p is used to refer to any other point(s), in general. Let the contour C be partitioned into two contour segments C1 (starting segment for tracing lines) and C2 (see Figure 4). Width profile was calculated using an intensity integrated Laplacian method6 which minimizes impact from irregular boundary of the chromosome segmentation by guiding the width profile trace lines to be contained within chromosome bands, which are regions with similar intensities. The width measurement of the normalized width profile at the discrete index λ (W (λ)) is obtained using the trace line which connects the contour points Cλ1 and Cλ2 from the two contours C1 and C2. Then, the set of candidate points for the centromere location pC (which stores the indices λ), where the local minima conditions of W(λ – 1) < W(λ) < W(λ + 1) and W(λ – 2) < W(λ) < W(λ + 2) are fulfilled for all valid locations λ of the width profile. In cases where the above condition failed to secure any candidates (mainly on extremely short chromosomes), the global minima was selected as the only candidate. Next, the following two sets of indices are created to correspond with each given element pC (α) of pC,\n\npmL(α) = W(β) where W(β) > W(γ), ∀γ < pC(α). Here pmL(α) stores the index of the global maxima for the portion (referred to as a regional maxima, henceforth) of the width profile prior to the candidate minima index pC(α).\n\npmR(α) = W(β) where W(β) > W(γ), ∀γ > pC(α). Similarly, pmR(α) stores the index of the global maxima for the portion of the width profile after the candidate minima index pC (α).\n\nThe width trace line, in red, connects the points Cλ1 and Cλ2 of the two contour segments.\n\nOnce the centromere candidate points pC and their corresponding maxima points pmL and pmR are calculated, the set of features Fc are calculated as given below. A set of 11 features F1c – F11c are proposed to train the third SVM classifier which will then be used to calculate the best candidate for a centromere location in a given chromosome. Features F1c to F3c provide an insight on the significance of the candidate point with respect to the general width profile distribution. The normalized width profile value itself is embedded in features F4c and F8c where the latter scales the minima based on the average value of the width profile. Features F5c and F6c capture the contour curvature values that are intrinsic to the constriction at the centromere location. Features F7c,F9c and F10c include distance measures which indicate the positioning of the candidate point with respect to the chromosome as well as to the width profile shape. Finally the feature F11c records the staining method used in the cell preparation. This gives the classifier a crucial piece of information that is then used to accommodate for specific shape features that may be the result of the particular laboratory procedure used to prepare and stain the sample.\n\nLet i be a candidate member number assigned among the pool of centromere candidates. Also, let d(1, i) be the Euclidean distance along the midpoints of the width profile trace lines (centerline) from a telomere to the candidate point, and L be the total length of the chromosome. In the following description, ∥.∥ represents the absolute value,\n\n1. F1c=‖ W(PC(i))−(PmL(i)) ‖. This feature calculates the absolute width profile difference between the candidate and the regional maxima prior to the candidate point on the width profile.\n\n2. F2c=‖ W(PC(i))−(PmR(i)) ‖. This feature calculates the absolute width profile difference between the candidate and the regional maxima beyond the candidate point on the width profile.\n\n3. F3c = F1c + F2c which calculates the combined width profile difference created by the candidate point.\n\n4. F4c = W(pC (i)). This captures the value of the width profile (0 ≤ F4c ≤ 1) at the candidate point location.\n\n5. F5c is the local curvature value at the contour point Cλ1 which corresponds to the current centromere candidate location (where λ = pC (i)).\n\n6. F6c is the local curvature value at the contour point Cλ2 which corresponds to the current centromere candidate location (where λ = pC (i)).\n\n7. F7c = min (d(1, i), L – d(1, i))/L. Gives a measure where the candidate is located with respect to the chromosome as a fractional measure (0 ≤ F7c ≤ 0.5).\n\n8. F8c = W(pC (i))/W– , where W– is the average of the width profile of the chromosome. This includes the significance of the candidate point minima with respect to the average width of the given chromosome.\n\n9. F9c = d(pmL(i), pC (i))/L. This gives the distance between the candidate point location and the regional maxima value prior to the candidate point, normalized by the total length of the chromosome.\n\n10. F10c = d(pC (i), pmR(i))/L. This gives the distance between the candidate point location and the regional maxima value beyond the candidate point, normalized by the total length of the chromosome.\n\n1. F11c is a Boolean feature used to indicate the staining process used during cell preparation. A value of ‘0’ would indicate the use of DAPI chromosome staining while ‘1’ would indicate a Giemsa-stained cell.\n\nThe detection of the centromere location assumes that each chromosome at least contains one centromere location within the chromosome. This is a reasonable assumption, since the centromere region is an integral part of chromosome anatomy which is normally retained in cell division, with the exception of acentric fragments produced by excessive radiation exposure, or rarely in congenital and neoplastic conditions. This assumption transforms the detection problem into a ranking problem in which we pick the best candidate from a pool of candidates. Therefore, this enables the same approach to be adopted that was utilized for the contour partitioning algorithm (section); i.e. in which the distance from the separating hyperplane (ρ) represents a measure of goodness-of-fit for a given candidate. This metric reduces the multidimensional feature space to a single dimension, which inherently reduces the complexity of the ranking procedure for the candidate locations. Since the large margin binary classifier (SVM) orients the separating hyperplane in the feature space, the 1D distance metric directly relates to how well a given candidate fits into the general characteristics of a given class label. A detailed introduction to the candidate-based centromere confidence metric is provided in the following section.\n\nAlthough existing measures of accuracy can establish performance of machine learning applications, these measures do not provide information on the reliability of the method. We developed a confidence metric for accurate detection of centromeres, which will be essential for assessment and ultimately adoption of this approach for diagnosis. We developed a Candidate Based Centromere Confidence metric (CBCC) to assess detection of a centromere location relative to alternatives. This value is obtained using the feature space derived via the classifier and the distance metric ρ. For a given set of candidate points, i.e. centromeres, of a chromosome pC, the goodness-of-fit (GF) of the optimal candidate point (ρ`) is obtained by calculating ‖(ρ `−ρ−)2‖, which is the average distance of all the remaining candidate points. In the ideal situation, the optimal candidate and the other candidates as support vectors for the classifier reside on opposite faces of the separating hyperplane (see Figure 5). Therefore the optimal candidate distance (ρ`) is ≈ 1, while the average of the remaining candidate distances (ρ–) is ≈ –1. The GF value is truncated at unity, since exceeding this value does not add additional information to the metric.\n\nThe blue square represents the optimal candidate while the other five candidates are given by the red squares in the feature space.\n\n\nResults\n\nThe complete data set used for developing and testing the algorithm discussed in this paper consists of 40 metaphase cell images, of which 38 consisted from irradiated samples obtained from cytogenetic biodosimetry laboratories and two were non-irradiated cells from a clinical cytogenetic laboratory. The chromosome data set comprised images of 18 Giemsa-stained cells and 22 DAPI-stained cells. The cells with minimal touching and overlapping chromosomes (a good metaphase spread) were manually selected from a pool of 1068 cell images for this experiment. Then 40 cell images were selected to represent both DAPI (55%) and Giemsa (45%) staining methods. During ground truth evaluation, the expert was presented with the set of centromere candidates generated by the algorithm and was asked to select the candidate that closely represented the correct chromosomal location, while explicitly marking other candidates as non-centromeres. In cases where all the candidates suggested by the algorithm were incorrect, all the positions were designated as negative candidates. Intra-observer variability between experts (ground truth) was minimal, as the laboratory directors differed in assessment in a single centromere out of > 500 chromosomes analyzed by both. The 1400 chromosome data set yielded 7058 centromere candidates. A randomly selected portion comprising 50% of this data set along with the corresponding ground truth centromere assignments were used for training a support vector machine for centromere localization. Next, the accuracy of centromere localization was calculated and is provided in Table 1. This provides a breakdown of the detection accuracy of the algorithm based on the presence or the absence of sister chromatid separation in the cell images for each staining method.\n\nTable 2 depicts CBCC values for accurately detected chromosomes as opposed to inaccurately detected chromosomes. It also includes a third category termed \"All nonviable candidate chromosomes\" (a subset of the inaccurate centromere detection category), where none of the candidates for a given chromosome were marked as capturing the true centromere of the chromosome.\n\nFigure 6 shows a representative sample of cases where the centromere was accurately localized. These cases include chromosomes with and without sister chromatid separation. The method does not detect centromere locations in all cases, some of which are impacted by the algorithm’s inability to fully correct for the adverse effects of sister chromatid separation (depicted in Figure 7).\n\nFigure 6(a) is a result from DAPI stained chromosomes while Figure 6(b)–(f) are results from Giemsa stained chromosomes. These results reported CBCC measures of (a) 1.000, (b) 1.000, (c) 1.000, (d) 0.995, (e) 1.000, (f) 0.661, respectively.\n\nThe detected centromere location (selected candidate) is depicted by a yellow dot while the segmented outline is drawn in blue. These results reported CBCC measures of (a) 0.368, (b) 0.066, (c) 0.655, respectively.\n\n\nDiscussion\n\nThe candidate based approach for centromere detection used a trained SVM classifier based on half of the input chromosomes. The accuracy of the method was then tested using the remaining 50% of the data set (2 fold cross validation); accuracy, sensitivity and specificity were 92%, 96% and 72%, respectively. Two fold cross validation was used instead of other methods such as the leave-one-out method, since it yields a reasonable estimation of the accuracy with a low computational cost. The higher sensitivity of this algorithm relative to our previous efforts5 can be attributed to improvements in the performance of the classifier on both typical and sister chromatid separated chromosomes. The lower specificity is predominantly related to lower confidence detection by the integrated intensity Laplacian algorithm of centromeres in acrocentric chromosomes, in which the centromeric constriction is not readily apparent because of its close proximity to one of the telomeres.\n\nThe objective of this study was to accurately detect the preferred centromere location (points) for each chromosome, even though the SVM produces a set of candidate points that can each be classified separately. All candidates in each chromosome were analyzed separately and the best candidate from this set was selected based on the distance metric value (ρ) of which the results are produced in Table 1. Upon testing, the algorithm accurately located a correct centromere location in 1220 of 1400 chromosomes (87%). This is a clear improvement on our previous attempt with an accuracy of 81% (detected centromere within 5 pixels of the known location) which used a much smaller dataset of 226 chromosomes. It is notable that 124 of the 180 chromosomes that were missed were instances of non-viable candidate chromosomes. Some of these were caused by segmentation of acrocentric chromosomes, where the lighter intensity of the short-arm satellite regions were segmented out, while others were primarily the result of an extreme degree of sister chromatid separation, such that the pairs of telomeres from sister chromatids could not be unequivocally paired. The values in Table 1 further suggest a slight reduction in accuracy for Giemsa-stained images, which contained significantly higher levels of sister chromatid separation and noisy chromosome boundaries.\n\nThe proposed method performed centromere localization accurately for chromosomes with high morphological variations (see Figure 6). From a machine learning point of view, Figure 6(a)–(c) are fairly straightforward centromere localizations. The CBCC values for all three cases were 1.000 which was truncated from an even higher value. This further validates the CBCC metric, indicating that the selected candidate is preferable over the other candidates in the same chromosome. It is important to notice that the boundary conditions at the telomeric region of Figure 6(c) is similar in appearance to those in Figure 3 or Figure 4. However, with further separation and intensity fading between the two sister chromatid arms, the segmentation algorithm could converge to a concave morphology in the telomere region that links the sister chromatids. Figure 6(e) represents such an instance where sister chromatid separation has had a significant effect on the chromosome segmentation. However, as a result of correcting for this effect, the algorithm has localized the centromere accurately with a CBCC value of 1.000. The chromosome segmentation in Figure 6(d) demonstrates evidence of extensive sister chromatid separation and therefore the CBCC value is at 0.995, which still is a high value for the data set. The Figure 6(f) represents a chromosome which is highly bent and also presents with significant sister chromatid separation. Nevertheless, the algorithm was capable of localizing an accurate centromere location though the CBCC value was low (0.661), which indicates a less than ideal separation among the centromere candidates.\n\nSome of the shortcomings of the proposed method are represented in Figure 7. Most of these (86%) were observed to be cases where none of the candidates were deemed to contain the actual centromere. This was mainly due to segmentation problems and add to high levels of sister chromatid separation. Figure 7(b) depicts an example where the segmentation algorithm failed to capture the constriction in an acrocentric chromosome. The CBCC value in this example was as low as 0.066, which indicates that the algorithm selected a weak candidate for the centromere. Figure 7(a) demonstrates a case where extreme sister chromatid separation has caused the segmentation algorithm to treat each individual chromatid separately. This chromosome had a low CBCC value of 0.368, which is consistent with the acentric nature (morphological) of the fragment. Figure 7(c) shows another impact of extreme sister chromatid separation on an acrocentric chromosome, namely, the incorrect connection of the long arm of a pair of sister chromatids, leading to an apparent, bent chromosome, instead of detecting sister chromatid separation. The CBCC measure fails to distinguish this chromosome from a normal bent chromosome, but nevertheless yielded a relatively high value of 0.655.\n\nAlthough not the focus of this study, we carried out a preliminary analysis of the capability of this algorithm to detect both centromeres in a set of dicentric chromosomes, which were present among an excess of normal single centromere chromomes, due to irradiation of some of the cytogenetic samples analyzed. The constriction at the second centromere is similar morphologically to the first centromere in these chromosomes, and therefore, it should be feasible that it be among the candidates found by the algorithm. We hypothesized that along with the optimal candidate, the second centromere was also expected to exhibit a short distance to the hyperplane and be well separated from the other candidates. These distances were compared for all centromere candidates, and probable dicentric chromosomes were identified by determining if the correct, ground truth centromeres were among the top four ranked candidates. The breakdown of the candidates which captured the second centromere location is given in Table 3, where 20 cases (out of 31) reported the second centromere location as the second highest ranked candidate location. Among the 31 dicentric chromosomes present in the data set, the first candidate (the selected centromere) was accurate in all instances. There were only two instances where the second centromere was not among the top four candidates. In both of these cases, the chromosomes exhibited a high degree of sister chromatid separation. Nevertheless, the proposed method provides a good framework for detecting dicentric chromosomes in radiation biodosimetry applications.\n\n\nConclusions\n\nWe have described a novel candidate-based centromere detection algorithm for analysis of metaphase cells prepared by different culturing and staining methods. The method performed with an 87% accuracy level when tested with a data set of 1400 chromosomes from a composite set of metaphase images. The algorithm was capable of correcting for the artifact created by premature sister chromatid separation. The majority of chromosomes with centromere constrictions were detected with very high sensitivity. We have also tested a promising extension of the centromere detection algorithm to accurately identify dicentric chromosomes for cytogenetic biodosimetry. Loss of specificity in both monocentric and dicentric chromosomes was the result of segmentation errors in acrocentric chromosomes, as well as in chromosomes with extreme degrees of sister chromatid separation.\n\nThe framework used for adding intensity into the Laplacian thickness measurement algorithm can be easily extended to include other features besides the calculation of chromosome width. Further investigation aimed at both improving centromere detection accuracy and applications of this algorithm to other detection problems is warranted. The Candidate Based Centromere Confidence (CBCC) was introduced as a measure for confidence in each centromere detection. However, this metric can be applied to any problem which requires a selection of a candidate from a pool of candidates. We suggest that the CBCC metric may be extensible to indicate the relative quality of a given cell image or of a set of meta-phase cells from the same patient. If successful, the CBCC metric may eventually limit the amount of time required to evaluate samples both prior to and during centromere detection.\n\n\nData and software availability\n\nZENODO: Chromosome images used for \"Centromere detection of human metaphase chromosome images using a candidate based method\", DOI: 10.5281/zenodo.5649017.\n\nZENODO: Matlab code for \"Centromere detection of human metaphase chromosome images using a candidate based method\", doi: 10.5281/zenodo.5649318.\n\nSource code license: GPL v3",
"appendix": "Author contributions\n\n\n\nAkila Subasinghe developed all the algorithms, performed manual selection of chromosomes from the data sets, performed all the tests and wrote the first draft of this paper. Jagath Samarabandu provided advice on selection of algorithms, proposed the key idea of using combinations of vertices for contour partitioning, helped select suitable features for classification and provided substantial amount of review, editing and incorporating reviewer feedback. Peter Rogan and Joan Knoll provided guidance in algorithm development, validation of results, and contributed to writing the paper. Ruth Wilkins and Farrah Flegal provided curated images of gamma irradiated metaphase cells and chromosomes.\n\n\nCompeting interests\n\n\n\nNone of the authors have any competing interests in any commercial entities or funding agencies mentioned in this article.\n\n\nGrant information\n\nSupported by the Western Innovation Fund (University of Western Ontario), Natural Sciences and Engineering Research Council of Canada and the DART-DOSE CMCR (5U01AI091173-02 from the US Public Health Service), the Canada Research Chairs Secretariat and the Canada Foundation for Innovation.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nArámbula Cosìo F, Vega L, Herrera Becerra A, et al.: Automatic identification of metaphase spreads and nuclei using neural networks. Med Biol Eng Comput. 2001; 39(3): 391–396. PubMed Abstract | Publisher Full Text\n\nMunot MV, Mukherjee J, Joshi M: A novel approach for efficient extrication of overlapping chromosomes in automated karyotyping. Med Biol Eng Comput. 2013; 51(12): 1325–1338. PubMed Abstract | Publisher Full Text\n\nMoradi M, Setarehdan SK, Ghaffari SR: Automatic locating the centromere on human chromosome pictures. In 16th IEEE Symposium on Computer-Based Medical Systems. 2003; 56–61. Publisher Full Text\n\nVaurijoux A, Gregoire E, Lefevre S, et al.: Detection of partial-body exposure to ionizing radiation by the automatic detection of dicentrics. Radiat Res. 2012; 178(4): 357–364. PubMed Abstract\n\nArachchige AS, Samarabandu J, Knoll JH, et al.: An image processing algorithm for accurate extraction of the centerline from human metaphase chromosomes. In International Conference on Image Processing (ICIP). 2010; 3613–3616. Publisher Full Text\n\nArachchige AS, Samarabandu J, Knoll JH, et al.: Intensity integrated laplacian-based thickness measurement for detecting human metaphase chromosome centromere location. IEEE Trans Biomed Eng. 2013; 60(7): 2005–2013. PubMed Abstract | Publisher Full Text\n\nMohammadi MR: Accurate localization of chromosome centromere based on concave points. J Med Signals Sens. 2012; 02(02): 88–94. PubMed Abstract | Free Full Text\n\nJahani S, Satarehdan SK: Centromere and length detection in artificially straightened highly curved human chromosomes. J Biol Eng. 2012; 02(5): 56–61. Publisher Full Text\n\nSchunck C, Johannes T, Varga D, et al.: New developments in automated cytogenetic imaging: unattended scoring of dicentric chromosomes, micronuclei, single cell gel electrophoresis, and fluorescence signals. Cytogenet Genome Res. 2004; 104(1–4): 383–389. PubMed Abstract | Publisher Full Text\n\nKobayashi T, et al.: Content and classification based ranking algorithm for metaphase chromosome images. In IEEE Conference on Multimedia Imaging, 2004.\n\nRogan PK, Li Y, Wickramasinghe A, et al.: Automating dicentric chromosome detection from cytogenetic biodosimetry data. Radiat Prot Dosimetry. 2014; 159(1–4): 95–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArachchige AS, Samarabandu J, Knoll JH, et al.: An accurate image processing algorithm for detecting fish probe locations relative to chromosome landmarks on dapi stained metaphase chromosome images. In Seventh Canadian Conference on Computer and Robot Vision (CRV), 2010; 223–230. Publisher Full Text\n\nBai X, Latecki LJ, Liu WY: Skeleton pruning by contour partitioning with discrete curve evolution. IEEE Trans Pattern Anal Mach Intel. 2007; 29(03): 449–62. PubMed Abstract | Publisher Full Text\n\nXu C, Kuipers B: Object detection using principal contour fragments. In Canadian Conference on Computer and Robot Vision (CRV), 2011; 363–370. Publisher Full Text\n\nLatecki LJ, Lakämper R: Polygon evolution by vertex deletion. In Proceedings of the Second International Conference on Scale-Space Theories in Computer Vision, Springer-Verlag London, UK, 1999; 1682: 398–409. Publisher Full Text\n\nStanley RJ, Keller J, Caldwell CW, et al.: Centromere attribute integration based chromosome polarity assignment. Proc AMIA Annu Fall Symp. 1996; 284–288. PubMed Abstract | Free Full Text\n\nSubasinghe A, Samarabandu J, et al.: Chromosome images used for \"Centromere detection of human metaphase chromosome images using a candidate based method\". Zenodo. 2016. Data Source\n\nSubasinghe A, Samarabandu J, et al.: Matlab code for \"Centromere detection of human metaphase chromosome images using a candidate based method\". Zenodo. 2016. Data Source"
}
|
[
{
"id": "14754",
"date": "13 Jul 2016",
"name": "Thomas Boudier",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a method to detect centromeres position in fluorescence images. The proposed method is an extension of their previous work where they segmented the chromosomes and detected the telomeres position. Based on the detected contour of the chromosome they extract salient points, and using am learning approach, try to infer the best position for the centromeres.\nThe methodology is described in details, but it is a bit difficult to follow the different steps, since there is no figures to illustrate the process. The simplified model of figure 1 should be extended with anchor points, telomeres, centromeres, so the different terms are clear for the reader.\nOne main point, however, is the usefulness of using machine learning, since the authors have first a set of 6 points, with 11 features each, and they want to determine the combination of the 6 points that describe best the chromosome. Since there are only 12 possible combinations, why not simply test them all and minimize some cost function ? The number of features used is also reduced, did the authors check the importance of each feature, using classical approaches like PCA ?\nFor the results, the authors should compare their new algorithm with other algorithms, or at least their own algorithm from previous work, to better emphasize the interest of this new method.\nFinally some minor comments :\nThe Giemsa staining should be referenced.\n\nFigure 5 is not useful.\n\nTypo euclidian.\n\nRephrase \"high resolution chromosomes\" (images)\n\nReferences to sections do not appear in the pdf file.",
"responses": []
},
{
"id": "14750",
"date": "14 Jul 2016",
"name": "Tanvi Arora",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI have read the research article, it is a good attempt made by the authors to present their work. It can been accepted for indexation, but before indexation the following points can be considered to revise the manuscript:\nKindly give citation for below:\n\"The reliable detection of the centromere by image analysis techniques is challenging due to the high morphological variations of chromosomes on microscope slides. This variation is caused by various cell preparation and staining methods along with other factors that occur during mitosis. Irregular boundaries and large variations in chromosome morphology can cause a detection algorithm to miss the constriction, especially in high resolution chromosomes.\"\n\nMissing information, like in the line just above methods\n\"The following section describes the proposed algorithm in detail. In section we show how this algorithm performed with a large data set and in section we comment on the performance and how it compares with other methods.\"\n\nThe authors have tested their method on DAPI & Q Banded metaspread images. But they have not taken the data from the standard dataset. I recommend them to test their method on standard dataset of ADIR dataset, Q Baded prometaphase dataset and G banded dataset. For which the benchmarked datasets are available online. Then compare their results on different datasets. As they have highlighted that straining methods can cause morphological variations.\n\nThe features have been selected for the purpose of classification. I would recommend that selected features should be analyzed using correlation based feature selection, to remove the redundant and non contributing features and improve the classification accuracy.\n\nThe result section can be further improved by explaining the reasons for obtaining such results.\n\nThere are grammatical errors.\n\nThis paper needs a second review.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1565
|
https://f1000research.com/articles/5-1564/v1
|
01 Jul 16
|
{
"type": "Case Report",
"title": "Case Report: 84 year-old woman with alien hand syndrome",
"authors": [
"Ihtesham Aatif Qureshi",
"Daniel Korya",
"Darine Kassar",
"Mohammed Moussavi",
"Daniel Korya",
"Darine Kassar",
"Mohammed Moussavi"
],
"abstract": "Background: Alien hand syndrome [AHS] is a rare and ill-defined neurological disorder. It produces complex, goal-directed motion of one hand that is involuntarily instigated. This syndrome characteristically arises after brain trauma, brain surgery, stroke or encephalitis. We describe a case of AHS in a patient who had a previous episode of subarachnoid hemorrhage affecting the left frontal lobe and corpus callosum. Case presentation: An 84-year-old woman presented to the emergency department complaining of headaches and several episodes of her left arm moving as if it was groping around trying to grab at her own body. A computed tomography scan of the head demonstrated an acute left superior frontal hemorrhage with compression of the corpus callosum. Transcranial Doppler report showed no significant abnormality in the insonated vessels. After being stabilized for the acute bleed, she was treated with clonazepam 0.5 mgat night for the uncontrolled hand movements. Her movements resolved by her next month follow up. The diagnosis of AHS was made based on her clinical presentation, characterization of the movement and localization correlating with findings in neuroimaging. Conclusion: We document a rare neurologic disorder seen in patients presenting with a history of previous strokes and a typical description of involuntary and unintentional, uncontrolled unilateral arm movements with repetitive grasping. The present case has a combination of frontal and callosal lesions. These findings appear to support a potential destruction leading to the rare syndrome.",
"keywords": [
"Alien-Hand Syndrome",
"Hemorrhagic Stroke",
"involuntary movements",
"headache"
],
"content": "Introduction\n\nAlien hand syndrome [AHS] is a loosely defined collection of symptoms characterized by involuntary movement of an upper limb in concurrence with the experience of separation from or exemplification of the actions of the limb itself. AHS was primarily used to label cases involving a disconnection of the hemispheres associated with a lesion of the corpus callosum1. There are three main types of AHS: frontal AHS (associated with lesions of the medial frontal cortex and characterized by reflex grasping and compulsive manipulation of tools); callosal AHS (associated with lesions of the corpus callosum and characterized by inter-manual conflict); and posterior cortical AHS (associated with parietal, occipital, thalamic structural lesions, cortico-basal degeneration, and characterized by sensory ataxia and feelings of estrangement of the upper limb)2,3.\n\nIn the present case, we describe a patient who presented with subarachnoid hemorrhage involving the left medial aspect (or parasagittal region) of the frontal lobe and caused the patient to present with headache and unconscious left hand movements consistent with AHS. We supplement the case presentation with a literature review of radiographic findings, past medical history and presenting symptoms of AHS.\n\n\nCase presentation\n\nAn 84-year-old African-American right hand dominant woman presented to the emergency department complaining of headache and episodes of uncontrollable left hand movements. She described the episodes wherein her left arm moved uncontrollably as if it was groping around trying to grab herself on her body. The patient explained that while asleep she felt that “someone is trying to grab meas if someone is in bed with me”. At times, she felt the need to talk to her hand or yell at it in order to command it to stop these embarrassing movements. These movements occurred while she was attempting to eat, watch television and during use of the toilet. The patient was evidently very distressed by these events and thought that her arm was “possessed by the Devil”.\n\nHer past medical history included chronic anemia, hypertension, previous intracranial hemorrhage, glaucoma, breast cancer (in remission) and cataracts. She was surgically treated for her breast cancer with lumpectomy and had bilateral cataract surgery in the past. Her family history consisted of diabetes, hypertension, migraine headaches and thyroid disease. She denied previous smoking or drinking. Her medications included anastrozole, latanoprost, amlodipine, pravastatin and multivitamin. She had no allergies.\n\nOn physical examination, her vitals were pulse 72/min, blood pressure was 140/70 mm Hg, weight was 130 lbs and height was 62 inches. She was comfortably seated and pleasant. No carotid bruits were appreciated. Heart rate and rhythm were regular. Normal pulses with no edema, cyanosis or clubbing noted on extremities. She was alert and oriented with fluent speech and no dysarthria. Cranial nerves II-XII were tested and normal. Fundi were benign with flat, well-marginated discs. There were no hemorrhages or exudates. The patient had intact visual fields to confrontation. Pupils were equal, round and reactive to light and accommodation. There was no afferent pupillary defect. Extra-ocular movements were normal. Facial motor and sensation were symmetric. Palate was symmetric on phonation and tongue protruded to the midline. Sternocleidomastoids were both full in strength.\n\nOn motor examination, she had strength of 5/5 on left arm and leg, whereas on the right side there was a considerable weakness with 2/5 strength on right arm and 3/5 on right leg. The deep tendon reflexes were brisk on the right. Plantar reflexes were down-going on the left and equivocal on the right. Coordination was worse on the left upper extremity with worse finger tapping. Sensory examination for pin-prick, temperature and vibration was symmetrically intact. The patient was able to stand and walk without assistance. Romberg test was negative.\n\nA computed tomography (CT) scan of the head was performed and showed an acute left frontal, parasagittal hematoma measuring 3 cm. Radiographic findings with magnetic resonance imaging (MRI) of the brain with and without contrast on axial MRI-T2 FLAIR (Figure 1) and sagittal MRI-T2 FLAIR (Figure 2) images were acquired.\n\nA Transcranial Doppler (TCD) study was done due to the clinical history of subarachnoid hemorrhage, and showed a normal direction of blood flow with mean flow velocities and spectral waveform within normal limits in all insonated segments of circle of Willis. The final impression was suggestive of no significant abnormality in the insonated vessels.\n\nThe patient was observed in the neurocritical care unit and then transferred to the stroke unit. After being stabilized, she was discharged to follow up within the neurovascular clinic. She presented to the clinic two weeks later with continued complaints of the left hand movements. At this point, the patient was expressing significant frustration and asked for assistance for these “Devilish movements”. After complete examination, the patient was recommended clonazepam 0.5 mg at night for the involuntary movements, was advised to avoid anti-platelet agents, and only take acetaminophen 500 mg for headaches. The patient was advised to return to the clinic after one month to determine if the medication was helping. At her one-month follow-up, the patient was pleased to report that her symptoms had decreased significantly and had last occurred about one week prior to her clinic visit.\n\n\nDiscussion\n\nAHS is a rare neurological disorder, typically seen during or after a vascular lesion of the brain. There are mainly three types of AHS. Two are frontal varieties, of which one is linked to lesions of the language dominant medial frontal cortex and the anterior corpus callosum affecting dominant hand, and the other involves the corpus callosum alone and affects the non-dominant hand4–7.The callosal-frontal alien hand variant usually results in more grasping activities and compulsive manipulations and the callosal AHS presents predominantly with inter-manual conflict4. The presenting case had a combination of the two frontal variants. Frontal variant lesion was caused by intracerebral and subarachnoid hemorrhage involving the territory of the anterior cerebral artery (ACA) affecting both the left frontal lobe and the corpus callosum.\n\nPremotor cortex, motor cortex and anterior two-thirds of the corpus callosum is supplied by ACA. Inter-hemispheric connections may be interrupted by occlusion of ACA. Proximal occlusions may result in motor weakness of the contralateral limbs along with compulsive reflex movements such as grasp reflex8.\n\nPatient with left hemispheric brain tumor invading the corpus callosum resulted in involuntary grasping and dropping of objects with the contralateral hand as described by Van Vleuten9. The term “alien hand syndrome” involves repetitive involuntary goal oriented limb movements acting opposite to the individual’s objective10. This type of movement was first reported by Van Vleuten.\n\nIn our case, the patient’s left hand was not only apraxic, but also accomplished distinctly improper actions, such as touching her right hand instead of nose, regardless of her understanding of the command, and failing to move when instructed.\n\nThe term “la main etrangere“ was coined by Brion and Jedynak11 describes failure to identify self-ownership of the limb or lack of self-care over the goal-directed limb movements among patients with callosal tumors. A milder version of this intermanual conflict was seen among patients with surgical callosal lesions. This was termed as “alien hand” by Bogen12, a translation of Brion and Jedynak’s la main etrangere.\n\nIn the case presented, the patient’s left arm was affected due to AHS, and the involuntary movement initially occurred daily, but later became less frequent. In a literature review, decrease in symptoms occurred in 68% of patients, whereas symptoms persisted in 32%13.\n\nOverall, the findings of this case represent frontal variants of previously described patterns causing AHS. This occurrence of the symptoms descriptive of the syndrome is seen after the affected corpus callosum’s communicating fibers disconnected from the functional left cortical region.\n\n\nConclusion\n\nThe development of AHS may be dependent on an injury pattern that first causes dysfunction of the association motor cortex of the right side and a subsequent lack of communication with the dominant left side, which would otherwise provide orders to the damaged non-dominant area. The patient’s symptoms were almost completely resolved with the use of clonazepam, but most patients improve spontaneously. As more reports emerge about this rare phenomenon, a greater understanding of the underlying mechanisms and causes of the dysfunction can be expected.\n\n\nPatient consent\n\nInformed written consent for publication of clinical details was obtained from the patient.",
"appendix": "Author contributions\n\n\n\nIAQ, DK have performed literature review and manuscript writing; DK, MM are involved in patient care and helped to make the diagnosis. All authors approved the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBundick T Jr, Spinella M: Subjective experience, involuntary movement, and posterior alien hand syndrome. J Neurol Neurosurg Psychiatry. 2000; 68(1): 83–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShereef H, Cavanna AE: The \"brother's arm:\" alien hand syndrome after right posterior parietal lesion. J Neuropsychiatry Clin Neurosci. 2013; 25(4): E02. PubMed Abstract | Publisher Full Text\n\nScepkowski LA, Cronin-Golomb A: The alien hand: cases, categorizations, and anatomical correlates. Behav Cogn Neurosci Rev. 2003; 2(4): 261–277. PubMed Abstract | Publisher Full Text\n\nFeinberg TE, Schindler RJ, Flanagan NG, et al.: Two alien hand syndromes. Neurology. 1992; 42(1): 19–24. PubMed Abstract | Publisher Full Text\n\nDenny-Brown D: Positive and negative aspects of cerebral cortical functions. N C Med J. 1956; 17(7): 295–303. PubMed Abstract\n\nGoldberg G, Bloom KK: The alien hand sign. Localization, lateralization and recovery. Am J Phys Med Rehabil. 1990; 69(5): 228–38. PubMed Abstract\n\nChan JL, Ross ED: Alien hand syndrome: influence of neglect on the clinical presentation of frontal and callosal variants. Cortex. 1977; 33(2): 287–99. PubMed Abstract | Publisher Full Text\n\nPark YW, Kim CH, Kim MO, et al.: Alien hand syndrome in stroke - case report & neurophysiologic study -. Ann Rehabil Med. 2012; 36(4): 556–560. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Vleuten CF: linksseitigemotorischeApraxie. Einbeitragzurphysiologie des balkens. AllgemeineZeitschrift fur Psychiatrie. 1907; 64: 203–39.\n\nGoldstein K: Zur Lehre von der motorischen Apraxie. J Psychol Neurol. 1908; 11: 169–87, 270–83. Reference Source\n\nBrion S, Jedynak CP: [Disorders of interhemispheric transfer (callosal disonnection). 3 cases of tumor of the corpus callosum. The strange hand sign]. Rev Neurol (Paris). 1972; 126: 257–66. PubMed Abstract\n\nBogen JE: The callosal syndrome. In: Heilman KM, Valenstein E, editors. Clinical neuropsychology. 1sted. New York: Oxford University Press, 1979; 308–59.\n\nKikkert MA, Ribbers GM, Koudstaal PJ: Alien hand syndrome in stroke: a report of 2 cases and review of the literature. Arch Phys Med Rehabil. 2006; 87(5): 728–732. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14782",
"date": "12 Jul 2016",
"name": "Nizar Souayah",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting case report of a patient who developed an AHS with involuntary movement after a stroke in the frontal and callosal area. The report was well written and supported by literature.",
"responses": []
},
{
"id": "16625",
"date": "10 Oct 2016",
"name": "Burak Yulug",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors well described an unusual case of AHS in CVD. However, I think that the following major comments would help to improve the impact of the manuscript:\nThe authors should clearly explain what they mean by stabilized. Are the AHS symptoms improved or do they mean the stabilization of neurovascular symptomatology only?\n\nAcute CT findings including the CT figure showing the acute cerebrovascular pathology should be added.\n\nThey should describe the frequency of similar combined cerebrovascular cases in the literature.\n\nWhat about the therapeutic effects of Clonazepam in the literature? This also should be added in the manuscript.\n\nIt is not clear how a left acute superior frontal pathology can lead to dysfunction of the right association cortex that secondary leads to a dysregulation of communication with the left side (Conclusion section). I think they should clearly define the hypothesized pathophysiology and include an illustration defining the disturbed communication and functional interaction in details.",
"responses": []
},
{
"id": "18081",
"date": "09 Dec 2016",
"name": "Mary Ann Thenganatt",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting report of a case of alien hand syndrome. It is clearly written and provides background on the history and classification of alien hand variants. The authors provide a detailed clinical description of this syndrome caused by a vascular insult, as well as discuss diagnostic studies, treatment and outcome of this case.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1564
|
https://f1000research.com/articles/5-1554/v1
|
30 Jun 16
|
{
"type": "Review",
"title": "Plant adaptation to drought stress",
"authors": [
"Supratim Basu",
"Venkategowda Ramegowda",
"Anuj Kumar",
"Andy Pereira",
"Supratim Basu",
"Venkategowda Ramegowda",
"Anuj Kumar"
],
"abstract": "Plants in their natural habitats adapt to drought stress in the environment through a variety of mechanisms, ranging from transient responses to low soil moisture to major survival mechanisms of escape by early flowering in absence of seasonal rainfall. However, crop plants selected by humans to yield products such as grain, vegetable, or fruit in favorable environments with high inputs of water and fertilizer are expected to yield an economic product in response to inputs. Crop plants selected for their economic yield need to survive drought stress through mechanisms that maintain crop yield. Studies on model plants for their survival under stress do not, therefore, always translate to yield of crop plants under stress, and different aspects of drought stress response need to be emphasized. The crop plant model rice (Oryza sativa) is used here as an example to highlight mechanisms and genes for adaptation of crop plants to drought stress.",
"keywords": [
"Adaptation",
"Drought tolerance",
"drought resistance",
"grain yield",
"rice",
"photosynthesis"
],
"content": "Introduction\n\nDrought stress is the most prevalent environmental factor limiting crop productivity1, and global climate change is increasing the frequency of severe drought conditions2. The sheer diversity of plant species grown across climatic regions that include extreme dry conditions suggests that, in nature, plants have evolved to endure drought stress with an array of morphological, physiological, and biochemical adaptations3. ‘Drought resistance’ (DR) is a broader term applied to plant species with adaptive features that enable them to escape, avoid, or tolerate drought stress4. ‘Drought escape’ is the ability of a plant species to complete its life cycle before the onset of drought. Thereby, plants do not experience drought stress, as they are able to modulate their vegetative and reproductive growth according to water availability, essentially through two different mechanisms: rapid phenological development and developmental plasticity5. Rapid phenological development involves rapid plant growth, producing a minimal number of seeds before the soil water depletes, and these plants are considered not to have any special morphological, physiological, or biochemical adaptations. Plants with mechanisms of developmental plasticity show little growth during the dry season, with very few flowers and seeds, but in wet seasons they grow indeterminately, producing a large amount of seed. ‘Drought avoidance’ is the ability of plants to maintain (relatively) higher tissue water content despite reduced water content in the soil4. This is achieved through a variety of adaptive traits involving the minimization of water loss (water savers) and optimization of water uptake (water spenders). Water spenders achieve higher tissue water status by maintaining the water uptake through increased rooting, hydraulic conductance, etc. under drought stress. In contrast, water savers use water effectively through reduced loss of water by reducing transpiration, transpiration area, radiation absorption, etc. under drought stress. ‘Drought tolerance’ (DT) is the ability of plants to endure low tissue water content through adaptive traits. These adaptive traits involve maintenance of cell turgor through osmotic adjustment and cellular elasticity, and increasing protoplasmic resistance6.\n\nImprovement of yield and maintaining yield stability of crops, under normal as well as drought stress conditions, is essential for the food security of the growing global population. It is difficult to resolve the role of different components of DR in the stability of the crop yield as the major objective. However, there exist a variety of different mechanisms for drought escape, avoidance, or tolerance in natural populations that can improve DR and maintain grain yield in crop plants. In nature, extreme DR is found in resurrection plants7,8 which possess strong drought escape mechanisms. Resurrection plants can be exposed to severe drought for months, extending up to years, forcing them to optimize their growth for survival, but not for seed production, in the long term9. Therefore, the DR mechanisms that enable plants to merely survive longer lead to subsistence yield, which is much lower than that which is observed under normal conditions. Crop plants, on the other hand, are grown by humans in environments under conditions for high agricultural production and will be exposed to only a random short-term drought stress of days to weeks, from which they must quickly respond to limit the damage caused by short-term drought stress while they continue to grow and yield in the stressful environments. Therefore, bringing in the drought adaptive mechanisms from plants adapted to grow in extreme dry conditions may not be a feasible option, as it may result in growth and/or yield penalty in crop plants under drought as well as normal conditions.\n\nAlthough plant survival is very critical in the early growth stages, the mechanisms have little relevance to increasing grain yield directly. The emphasis to improve DR of crop plants should therefore be based on stability of yield components and not on plant survival alone. So far, most of the efforts to improve grain yield under drought stress were focused on secondary traits such as root architecture, leaf water potential, osmotic adjustment, and relative water content at the vegetative stage, which are often not highly correlated with grain yield10,11. Looking forward in crops, the effective drought improvement approach should be selection for yield and its component traits under reproductive-stage drought stress12. Additionally, little importance has also been given to simultaneous improvement of grain yield under normal and drought conditions. Selection for DT has been suggested to have a yield drag under normal conditions. It has been proposed that the yield potential of crop plants should be simultaneously selected for under favorable and environmental stress conditions, as there is a positive correlation between yield potential under normal and drought stress conditions13. Combining high yield potential under normal conditions with good yield under drought stress is the ideal trait. Identification of mechanisms, traits, and genes regulating yield under drought stress that are free from yield drag under normal conditions should be the focus. For example, regulation of yield under normal as well as drought stress conditions has been shown for three NAC family transcription factors (TFs). Transgenic plants expressing OsNAC5, OsNAC9, and OsNAC10 TFs showed an increase in grain yield of 5-26% under normal conditions14–16. Nevertheless, in these studies, yield under normal conditions has been overlooked with more emphasis given to yield under drought stress. Two of our recent studies show the potential of simultaneously improving and stabilizing grain yield, both under normal as well as drought stress conditions, using two regulatory genes, namely GUDK and HYR in rice17,18. These studies indicate that it might be advantageous to identify mechanisms and genes for increasing grain yield that are also stable or maintained under drought stress conditions.\n\nDespite the complexity of DR, tremendous progress has been made in understanding the drought-adaptive mechanisms of plants1,19,20. Adaptation through DR mainly involves morpho-physiological alterations. These alterations in adaptive processes are controlled by molecular mechanisms that regulate the expression of genes21. There exists a large diversity in drought adaptation within a crop species, as some genotypes are able to cope with drought better than others. Genotypes that differ in drought adaptive mechanisms serve as an important resource to study the variation in drought adaption in crop plants. This natural variation needs to be exploited to simultaneously improve DR and yields of cultivated varieties through better understanding of the underlying mechanisms and to aid in selection for these traits22. In the following sections, we describe the widely known morpho-physiological processes and recent molecular advances in regulating these drought-adaptive processes leading to increased yield in crop plants.\n\n\nPhotosynthesis\n\nDrought stress is known to reduce photosynthesis by decreasing both leaf area and photosynthetic rate per unit leaf area. Reduced photosynthetic rate is mainly through stomatal closure or metabolic impairment23. Continued photosynthetic light reactions during drought stress under limited intercellular CO2 concentration results in the accumulation of reduced photosynthetic electron transport components, which can potentially reduce molecular oxygen, resulting in the production of reactive oxygen species (ROS). ROS can cause severe damage to the photosynthetic apparatus24. The adaptive responses that plants have developed to reduce drought-induced damage to photosynthesis include thermal dissipation of light energy, the xanthophyll cycle, the water-water cycle, and dissociation of the light-harvesting complexes from photosynthetic reaction centers25–27. The metabolic impairment during drought stress is mainly caused by changes in photosynthetic carbon metabolism24. The biochemical efficiency of photosynthesis under drought stress mainly depends on ribulose-1,5-bisphosphate (RuBP) regeneration and the activity of ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO)28,29. Considerable progress has been made in improving the stomatal components for CO2 diffusion, photosynthetic light reaction, and metabolic changes, including the expression of photosynthesis-related genes to regulate photosynthesis under drought towards the improvement of grain yield30.\n\nThe C4 pathway of carbon assimilation has been suggested to be the major adaptation of the C3 pathway to limit water loss, reduce photorespiration, and improve photosynthetic efficiency under drought stress31. However, many important crops—including rice, wheat, soybean, and potato—use the C3 pathway of photosynthesis. Although the transfer of the C4 pathway into C3 crops is underway, so far its contribution to increased grain yield is very limited32. Photosynthetic adaptation of plants to drought stress involves a complex interaction of hormones, ROS, sugars, and other metabolic events33. Combinations of computational models, which integrate the physiological and metabolic processes with gene expression data, along with modern breeding and transgenic technologies hold promise in improving photosynthesis and hence crop yield under normal as well as drought stress conditions.\n\nIn recent studies, we used a rice gene regulatory network to identify a TF termed HYR (HIGHER YIELD RICE), which was highly associated with primary carbon metabolism17, and on overexpression in rice enhanced photosynthesis under normal conditions as well as under drought and high temperature stress. HYR regulates several morpho-physiological processes leading to higher yield under normal and environmental stress conditions. Our study showed that HYR is a master regulator of photosynthesis, directly activating photosynthesis genes, cascades of TFs, and other downstream genes involved in photosynthetic carbon metabolism, resulting in improved yield.\n\n\nHormonal regulation\n\nMajor phytohormones, such as abscisic acid (ABA), cytokinin (CK), gibberellic acid (GA), auxin, and ethylene, regulate diverse processes which enable plant adaptation to drought stress34. Upon exposure of plants to drought stress, ABA is the major hormone synthesized in roots and translocated to leaves to initiate adaptation of plants to drought stress through stomatal closure and reduced plant growth35. However, modulating the ABA-induced drought adaptation of plants for better yield remains a greater challenge because of the potential inadvertent reduction in carbon gain upon stomatal closure and ABA-induced senescence, especially if the drought occurs at the reproductive stage36. There are ABA signaling genes, such as OsNAP, OsNAC5, and DSM2, which promote improved yield under reproductive drought37–40. These ABA-induced non-stomatal adaptations of plants under drought stress can be exploited to improve grain yield under reproductive drought (Figure 1).\n\nThis pathway is required for drought stress tolerance and grain yield under drought. The drought stress ABA-dependent signal is shown perceived directly by the regulatory genes described in the text, followed by transcriptional regulation of downstream genes and underlying stress response mechanisms. Genes regulating drought tolerance (DT) at the vegetative stage are shaded green, and genes regulating DT and grain yield under drought are shaded orange. The resulting phenotypes are represented for DT at the vegetative level only (green diamonds), or DT and grain yield (orange diamonds). The genes described here are OsGH392, OsNAP, OsABI239, AP3793, OsPP2C09, OsPP2C0694, OsPYL/RCAR5, OsSIDP36695, OsMYB4896, OsRK197, Oshox2298, SNAC299,100, OsOAT101, OsbZIP23102, SNAC199, OsEREBP1103, OsbZIP71104, OsbZIP46105, OsABI5106, DSM240, AREB2107, OsSRO1c108, and OsABA8OX3109.\n\nUnder drought stress, CKs are known to delay premature leaf senescence and death, adaptive traits very useful for increasing grain yield. An increase in the endogenous levels of CK through expression of isopentenyltransferase (IPT), a CK biosynthetic pathway gene, leads to stress adaptation by delaying drought-induced senescence and an increase in yield41,42. Generally, auxin has been shown to negatively regulate drought adaptation in plants. Decrease in indole-3-acetic acid (IAA) content was shown to be associated with up-regulation of genes encoding late embryogenesis abundant (LEA) proteins, leading to drought adaptation in plants43,44. Recently, the DEEPER ROOTING 1 (DRO1) gene determining a quantitative trait locus (QTL) controlling root growth angle was shown to be negatively regulated by auxin. Higher expression of DRO1 in a shallow-rooting rice cultivar resulted in drought avoidance and high yield under drought45. GA is suggested to positively regulate plant adaptation to drought stress. A rapid decline in levels of endogenous GA was observed in plants subjected to drought stress, resulting in growth inhibition46. The role of GA in regulating grain yield of crop plants is thus an important area that can be further explored. Ethylene is a negative regulator of drought stress response by promoting leaf senescence and inhibiting root growth and development, shoot/leaf expansion, and photosynthesis47–51. Ethylene can also directly affect yield by increasing embryo and grain abortion and reducing the grain-filling rate52. In addition to the major hormones, other hormones such as brassinosteroids, jasmonic acid (JA), salicylic acid (SA), and strigolactone also have an equally important role in plant growth and development. However, their function under drought stress is relatively less characterized. Tillering in rice has been suggested to be the outcome of an interaction among three hormones, CK, auxin, and strigolactone, with CK promoting branching and the other two inhibiting it53,54, suggesting that all hormones do not act in isolation but instead interact and modulate each other’s biosynthesis and responses. Therefore, the net outcome of the drought stress response is regulated by a balance between hormones that promote and those that inhibit the trait, rather than individual hormones.\n\n\nTranspiration and stomatal conductance\n\nThe immediate response of plants on being exposed to drought stress is stomatal closure. However, stomatal closure not only diminishes water loss through transpiration but also reduces CO2 and nutrient uptake, and hence alters metabolic pathways such as photosynthesis55. Plants growing in dry areas have developed xeromorphic traits to reduce transpiration under drought stress. Reduction in transpiration under drought stress conditions can also be achieved through leaf shedding (i.e. deciduous species in drought) as well as decrease in leaf number, leaf size, and branching. Another adaptation to counter drought stress is sclerophylly, where plants form hard leaves that will not suffer permanent damage due to wilting and can be restored to full functionality when normal conditions resume56. Recent research has shown that decreased stomatal conductance in response to drought stress is related not only to reduced expression of aquaporin genes but also to anatomical traits leading to reduction of chloroplast surface area exposed to intercellular space per unit leaf area57,58. Several other factors, including leaf developmental stage and light availability, are also known to interact with drought in modulating mesophyll and chloroplast differentiation, ultimately affecting conductance and photosynthetic capacity58. Reduction in stomatal size and number on exposure to drought is another adaptation for survival under drought conditions. Previous studies have shown that while there is an increase in stomatal density under mild drought stress, there is a decrease during severe drought59. Thus, all these adaptations in plants reduce the negative impacts of drought stress on photosynthesis and thereby have a positive effect on water use efficiency (WUE), which in turn will result in high yield potential and high yield60. Such an adaptation was shown in rice by overexpression of the Arabidopsis AP2/ERF TF HARDY that improved WUE (the ratio of biomass produced to water used) by enhancing photosynthesis and reducing transpiration61. The above reported traits therefore exemplify adaptive mechanisms in plants to survive under drought stress without loss of productivity or yield.\n\n\nRoot morphology\n\nIn many agriculturally important crops, drought stress is perceived first by the root system, which continues to grow underneath the soil even though shoot growth is inhibited under these conditions62. Although the growth of the primary root is not affected by drought stress, the growth of lateral roots is significantly reduced, mainly by suppression of the activation of the lateral root meristems63. The Arabidopsis R2R3-type MYB TF MYB96 has been shown to regulate activation of lateral root meristem through an ABA signaling cascade, with an activation-tagged mutant showing enhanced DR with reduced lateral root formation64. The plant microRNA miR393 has also been shown to play a role in root-mediated adaptation to drought stress response through attenuation of auxin signaling65. In addition to the lateral roots, the presence of small roots is also considered as an adaptive strategy to increase water uptake by providing more absorptive surface. Presence of specialized tissues like rhizodermis, with a thickened outer cell wall or suberized exodermis, or reduction in the number of cortical layers are considered an adaptive advantage for drought stress survival. Hydrotropism is another adaptive measure taken by plants to counter stress, where studies have shown that degradation of amyloplasts in the columella cells of plant roots on exposure to drought stress increases hydrotropism66,67. Hormonal cross-talk mediated by auxin, CK, GA, and ABA has been implicated as a potential chemical signal in response to water stress to modulate root system architecture68.\n\nThe expression of enzymes related to root morphology (e.g. xyloglucan endotransglucosylase) is induced upon mild drought stress, while other structural proteins are down-regulated, which is strongly correlated with root growth and hence an augmentation in the surface area for water uptake. The alterations in the expression of these proteins correlate positively with lateral development that in turn also affects photosynthesis69. More lateral root and root hair formation was found in lines possessing a QTL, qDTY12.1, only when under drought70. Such traits, which are expressed only under drought stress, have higher potential to increase grain yield under drought. Moreover, it has also been shown that drought stress triggers a wide variety of anatomical traits expressed to different levels and patterns in different species and even in different cultivars within species71–73. For example, suberization and compaction of sclerenchyma layer cells were shown to decrease in rice under drought, which increases retention of water under drought stress71.\n\n\nOsmotic adjustment\n\nOsmotic adjustment (OA) is defined as a process of solute accumulation in dividing cells when the water potential is reduced, and thereby helps in maintaining the turgor74. Cell enlargement and growth in plants is highly dependent on water availability and helps in maintaining the turgor. Turgor measurement in growing regions of plants, especially the leaves and stems, shows little or no reduction, though cell enlargement is inhibited during drought stress and is believed to be due to OA75,76. Under conditions of drought stress, OA has been implicated in maintaining stomatal conductance, photosynthesis, leaf water volume, and growth74,77. At times of drought stress, in addition to the reduction in water content, there are also other associated changes, such as increases in salt concentration and mechanical impedance78. Inorganic cations, organic acids, carbohydrates, and free amino acids are the known predominant solutes that accumulate in response to water stress. Previous studies have shown that drought-resistant wheat varieties, with yield stability under drought stress, have a greater capacity for osmoregulation than less resistant varieties76. The accumulation of compatible solutes such as proline and glycine betaine help in protecting the plants from detrimental effects of drought stress not only by OA but also by detoxification of ROS, protection of membrane integrity, and stabilization of enzymes or proteins79. Enzymes such as betaine aldehyde dehydrogenase (BADH), pyrroline-5-carboxylate reductase (P5CR), and ornithine δ-aminotransferase (OAT) have been shown to play major roles in OA. Overexpression of Arabidopsis EDT1/HDG11 was shown to increase DT of poplar and cotton through increased accumulation of solutes such as proline and soluble sugars and also increase the yield of cotton in the field80. However, there are some plants in which sugars are the main osmolytes that play a significant role in OA, including sucrose, trehalose, glucose, and fructose. Previous studies have shown that overexpression of the sucrose:fructan-6-fructosyltransferase (6-SFT) gene from Psathyrostachys huashanica in tobacco and the trehalose-6-phosphate phosphatase gene OsTPP1 in rice confers abiotic stress tolerance81,82. Researchers have also identified a QTL for OA on chromosome 8 in rice that is homeologous with a segment of wheat chromosome 783.\n\n\nSource-sink relationships\n\nSource-sink relationships largely determine the grain yield of cereal crops, with developing grains being primary sinks while the top two leaves, the flag leaf in particular, are the primary source84,85. Drought stress affects the source-sink relationship by reducing the source strength, leading to yield reduction. Sufficient sugar supply through photosynthesis, transport, and conversion of sugars is regarded as the most critical component in determining the viability of reproductive organs in rice86–88. Drought stress dramatically affects pollen viability due to abnormal starch accumulation89. Insufficient starch synthesis and arrested pollen development have been linked to reduced invertase activity under drought stress88. In addition to invertases, active grain filling involves other key enzymes, such as sucrose synthase, ADP glucose pyrophosphorylase (AGPase), and starch synthase as well as starch branching and debranching enzymes, which are also affected by drought stress88. Therefore, enhancing rice yield through source-sink relationships involves not only sugar metabolism but also the regulated mobilization of metabolic resources from source to sink tissue.\n\n\nFuture research perspectives\n\nImproving DR in crop plants is a challenge for plant breeders and crop physiologists, as it is a complex genetic trait with multiple pathways involved. Effective development of drought-resistant crop plants thus requires the pyramiding and interaction of many mechanisms, traits, and genes that are appropriate to individual crops and their growing environments. Success in this direction not only extends the growing area of crop plants but also achieves stable yield in drought-prone areas. Identifying genetic variation for DR is the first step towards development of drought-resistant crop plants. Such variation is often present in wild species and adapted genotypes that have evolved under natural selection and these are the best source of DR traits. Evaluation of these resources through an integrated phenotyping and genotyping approach under field conditions alongside identification of traits that are directly associated with yield is key to improving DR90. Comparative omics analyses between the diverse germplasm could aid in bettering our understanding of the variety of crop adaptations to drought stress. In addition, analysis of specific genes focussed on increasing DR while stabilizing the yield is crucial for understanding the broad basis of complex traits such as DR91. The development of high-yielding resilient crops74 that maintain yield stability under drought and other environmental stresses due to climate change is also currently needed. Drought-resistant plants should combine a better root system, stomatal regulation, WUE, and hormonal balance while avoiding the negative effects on grain yield under both normal and drought stress conditions. Therefore, a holistic crop improvement strategy should involve the deployment of high crop yield potential and the utilization of a combination of morpho-physiological, biochemical, and anatomical adaptive responses to drought stress.",
"appendix": "Author contributions\n\n\n\nAll authors contributed to writing the manuscript. Supratim Basu and Andy Pereira prepared Figure 1.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work was supported by the National Science Foundation award numbers DBI–0922747 and ABI1062472.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBray EA: Plant responses to water deficit. Trends Plant Sci. 1997; 2(2): 48–54. Publisher Full Text\n\nDai A: Increasing drought under global warming in observations and models. Nat Clim Chang. 2012; 3: 52–8. Publisher Full Text\n\nBohnert HJ, Nelson DE, Jensen RG: Adaptations to Environmental Stresses. Plant Cell. 1995; 7(7): 1099–111. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevitt J: Responses of plants to environmental stresses. Volume II. Water, radiation, salt, and other stresses. (Academic Press), 1980. Reference Source\n\nJones MM, Turner NC, Osmond CB: Mechanisms of drought resistance. The physiology and biochemistry of drought resistance in plants. 1981; 1: 15–37.\n\nMorgan JM: Osmoregulation and Water Stress in Higher Plants. Annu Rev Plant Physiol. 1984; 35: 299–319. Publisher Full Text\n\nGaff DF: Desiccation tolerant plants in South America. Oecologia. 1987; 74(1): 133–6. Publisher Full Text\n\nGaff DF: Desiccation-tolerant flowering plants in southern Africa. Science. 1971; 174(4013): 1033–4. PubMed Abstract | Publisher Full Text\n\nPorembski S: Granitic and gneissic outcrops (inselbergs) as centers of diversity for desiccation-tolerant vascular plants. Plant Ecol. 2000; 151(1): 19–28. Publisher Full Text\n\nJongdee B, Fukai S, Cooper M: Leaf water potential and osmotic adjustment as physiological traits to improve drought tolerance in rice. Field Crops Res. 2002; 76(2–3): 153–63. Publisher Full Text\n\nPantuwan G, Fukai S, Cooper M, et al.: Yield response of rice (Oryza sativa L.) genotypes to drought under rainfed lowland: 3. Plant factors contributing to drought resistance. Field Crops Res. 2002; 73(2–3): 181–200. Publisher Full Text\n\nVenuprasad R, Lafitte HR, Atlin GN: Response to Direct Selection for Grain Yield under Drought Stress in Rice. Crop Sci. 2007; 47(1): 285–293. Publisher Full Text | Faculty Opinions Recommendation\n\nGuan YS, Serraj R, Liu SH, et al.: Simultaneously improving yield under drought stress and non-stress conditions: a case study of rice (Oryza sativa L.). J Exp Bot. 2010; 61(15): 4145–56. PubMed Abstract | Publisher Full Text\n\nJeong JS, Kim YS, Redillas MC, et al.: OsNAC5 overexpression enlarges root diameter in rice plants leading to enhanced drought tolerance and increased grain yield in the field. Plant Biotechnol J. 2013; 11(1): 101–14. PubMed Abstract | Publisher Full Text\n\nRedillas MC, Jeong JS, Kim YS, et al.: The overexpression of OsNAC9 alters the root architecture of rice plants enhancing drought resistance and grain yield under field conditions. Plant Biotechnol J. 2012; 10(7): 792–805. PubMed Abstract | Publisher Full Text\n\nJeong JS, Kim YS, Baek KH, et al.: Root-specific expression of OsNAC10 improves drought tolerance and grain yield in rice under field drought conditions. Plant Physiol. 2010; 153(1): 185–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmbavaram MM, Basu S, Krishnan A, et al.: Coordinated regulation of photosynthesis in rice increases yield and tolerance to environmental stress. Nat Commun. 2014; 5: 5302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamegowda V, Basu S, Krishnan A, et al.: Rice GROWTH UNDER DROUGHT KINASE is required for drought tolerance and grain yield under normal and drought stress conditions. Plant Physiol. 2014; 166(3): 1634–45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYordanov I, Velikova V, Tsonev T: Plant responses to drought, acclimation, and stress tolerance. Photosynthetica. 2000; 38(2): 171–86. Publisher Full Text\n\nBray EA: Molecular Responses to Water Deficit. Plant Physiol. 1993; 103(4): 1035–40. PubMed Abstract | Free Full Text\n\nOsakabe Y, Osakabe K, Shinozaki K, et al.: Response of plants to water stress. Front Plant Sci. 2014; 5: 86. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nReynolds M, Tuberosa R: Translational research impacting on crop productivity in drought-prone environments. Curr Opin Plant Biol. 2008; 11(2): 171–9. PubMed Abstract | Publisher Full Text\n\nTezara W, Mitchell VJ, Driscoll SD, et al.: Water stress inhibits plant photosynthesis by decreasing coupling factor and ATP. Nature. 1999; 401: 914–7. Publisher Full Text\n\nLawlor DW, Cornic G: Photosynthetic carbon assimilation and associated metabolism in relation to water deficits in higher plants. Plant Cell Environ. 2002; 25(2): 275–94. PubMed Abstract | Publisher Full Text\n\nNiyogi KK: PHOTOPROTECTION REVISITED: Genetic and Molecular Approaches. Annu Rev Plant Physiol Plant Mol Biol. 1999; 50: 333–59. PubMed Abstract | Publisher Full Text\n\nDemmig-Adams B, Adams WW III: The role of xanthophyll cycle carotenoids in the protection of photosynthesis. Trends Plant Sci. 1996; 1(1): 21–6. Publisher Full Text\n\nDemmig-Adams B, Adams WW 3rd: Photoprotection in an ecological context: the remarkable complexity of thermal energy dissipation. New Phytol. 2006; 172(1): 11–21. PubMed Abstract | Publisher Full Text\n\nLawlor DW: Limitation to photosynthesis in water-stressed leaves: stomata vs. metabolism and the role of ATP. Ann Bot. 2002; 89(Spec No): 871–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMedrano H, Parry MAJ, Socias X, et al.: Long term water stress inactivates Rubisco in subterranean clover. Ann Appl Biol. 1997; 131(3): 491–501. Publisher Full Text\n\nChaves MM, Flexas J, Pinheiro C: Photosynthesis under drought and salt stress: regulation mechanisms from whole plant to cell. Ann Bot. 2009; 103(4): 551–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards G, Walker D: C3, C4: mechanisms, and cellular and environmental regulation, of photosynthesis. Univ of California Press, 1983. Reference Source\n\nGowik U, Westhoff P: The path from C3 to C4 photosynthesis. Plant Physiol. 2011; 155(1): 56–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinheiro C, Chaves MM: Photosynthesis and drought: can we make metabolic connections from available data? J Exp Bot. 2011; 62(3): 869–82. PubMed Abstract | Publisher Full Text\n\nWilkinson S, Kudoyarova GR, Veselov DS, et al.: Plant hormone interactions: innovative targets for crop breeding and management. J Exp Bot. 2012; 63(9): 3499–509. PubMed Abstract | Publisher Full Text\n\nWilkinson S, Davies WJ: Drought, ozone, ABA and ethylene: new insights from cell to plant to community. Plant Cell Environ. 2010; 33(4): 510–25. PubMed Abstract | Publisher Full Text\n\nJi X, Dong B, Shiran B, et al.: Control of abscisic acid catabolism and abscisic acid homeostasis is important for reproductive stage stress tolerance in cereals. Plant Physiol. 2011; 156(2): 647–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSperotto RA, Ricachenevsky FK, Duarte GL, et al.: Identification of up-regulated genes in flag leaves during rice grain filling and characterization of OsNAC5, a new ABA-dependent transcription factor. Planta. 2009; 230(5): 985–1002. PubMed Abstract | Publisher Full Text\n\nLiang C, Wang Y, Zhu Y, et al.: OsNAP connects abscisic acid and leaf senescence by fine-tuning abscisic acid biosynthesis and directly targeting senescence-associated genes in rice. Proc Natl Acad Sci U S A. 2014; 111(27): 10013–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChen X, Wang Y, Lv B, et al.: The NAC family transcription factor OsNAP confers abiotic stress response through the ABA pathway. Plant Cell Physiol. 2014; 55(3): 604–19. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nDu H, Wang N, Cui F, et al.: Characterization of the beta-carotene hydroxylase gene DSM2 conferring drought and oxidative stress resistance by increasing xanthophylls and abscisic acid synthesis in rice. Plant Physiol. 2010; 154(3): 1304–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeleg Z, Reguera M, Tumimbang E, et al.: Cytokinin-mediated source/sink modifications improve drought tolerance and increase grain yield in rice under water-stress. Plant Biotechnol J. 2011; 9(7): 747–58. PubMed Abstract | Publisher Full Text\n\nPeleg Z, Blumwald E: Hormone balance and abiotic stress tolerance in crop plants. Curr Opin Plant Biol. 2011; 14(3): 290–5. PubMed Abstract | Publisher Full Text\n\nXie Z, Jiang D, Cao W, et al.: Relationships of endogenous plant hormones to accumulation of grain protein and starch in winter wheat under different post-anthesis soil water statusses. Plant Growth Regul. 2003; 41(2): 117–27. Publisher Full Text\n\nZhang S, Li C, Cao J, et al.: Altered architecture and enhanced drought tolerance in rice via the down-regulation of indole-3-acetic acid by TLD1/OsGH3.13 activation. Plant Physiol. 2009; 151(4): 1889–901. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUga Y, Sugimoto K, Ogawa S, et al.: Control of root system architecture by DEEPER ROOTING 1 increases rice yield under drought conditions. Nat Genet. 2013; 45(9): 1097–102. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWang C, Yang A, Yin H, et al.: Influence of water stress on endogenous hormone contents and cell damage of maize seedlings. J Integr Plant Biol. 2008; 50(4): 427–34. PubMed Abstract | Publisher Full Text\n\nMunné-Bosch S, Alegre L: Die and let live:Leaf senescence contributes to plant survival under drought stress. Funct Plant Biol. 2004; 31(3): 203–216. Publisher Full Text\n\nFukao T, Xu K, Ronald PC, et al.: A variable cluster of ethylene response factor-like genes regulates metabolic and developmental acclimation responses to submergence in rice. Plant Cell. 2006; 18(8): 2021–34. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPerata P, Voesenek LA: Submergence tolerance in rice requires Sub1A, an ethylene-response-factor-like gene. Trends Plant Sci. 2007; 12(2): 43–6. PubMed Abstract | Publisher Full Text\n\nRajala A, Peltonen-Sainio P: Plant Growth Regulator Effects on Spring Cereal Root and Shoot Growth. Agron J. 2001; 93(4): 936–943. Publisher Full Text\n\nSharp RE: Interaction with ethylene: changing views on the role of abscisic acid in root and shoot growth responses to water stress. Plant Cell Environ. 2002; 25(2): 211–22. PubMed Abstract | Publisher Full Text\n\nYang JC, Zhang JH, Ye YX, et al.: Involvement of abscisic acid and ethylene in the responses of rice grains to water stress during filling. Plant Cell Environ. 2004; 27(8): 1055–64. Publisher Full Text\n\nGomez-Roldan V, Fermas S, Brewer PB, et al.: Strigolactone inhibition of shoot branching. Nature. 2008; 455(7210): 189–94. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nXing Y, Zhang Q: Genetic and molecular bases of rice yield. Annu Rev Plant Biol. 2010; 61: 421–42. PubMed Abstract | Publisher Full Text\n\nXiong L, Zhu J: Molecular and genetic aspects of plant responses to osmotic stress. Plant Cell Environ. 2002; 25(2): 131–9. PubMed Abstract | Publisher Full Text\n\nMicco VD, Aronne G: in Plant Responses to Drought Stress. (ed Aroca. Ricardo). Springer, 2002.\n\nMiyazawa S, Yoshimura S, Shinzaki Y, et al.: Deactivation of aquaporins decreases internal conductance to CO2 diffusion in tobacco leaves grown under long-term drought. Funct Plant Biol. 2008; 35(7): 553–564. Publisher Full Text\n\nTosens T, Niinemets U, Vislap V, et al.: Developmental changes in mesophyll diffusion conductance and photosynthetic capacity under different light and water availabilities in Populus tremula: how structure constrains function. Plant Cell Environ. 2012; 35(5): 839–56. PubMed Abstract | Publisher Full Text\n\nXu Z, Zhou G: Responses of leaf stomatal density to water status and its relationship with photosynthesis in a grass. J Exp Bot. 2008; 59(12): 3317–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlum A: Drought resistance, water-use efficiency, and yield potential—are they compatible, dissonant, or mutually exclusive? Aust J Agric Res. 2005; 56(11): 1159–1168. Publisher Full Text\n\nKaraba A, Dixit S, Greco R, et al.: Improvement of water use efficiency in rice by expression of HARDY, an Arabidopsis drought and salt tolerance gene. Proc Natl Acad Sci U S A. 2007; 104(39): 15270–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpollen WG, Sharp RE: Spatial distribution of turgor and root growth at low water potentials. Plant Physiol. 1991; 96(2): 438–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeak KI, Malamy J: Osmotic regulation of root system architecture. Plant J. 2005; 43(1): 17–28. PubMed Abstract | Publisher Full Text\n\nSeo PJ, Park C: Auxin homeostasis during lateral root development under drought condition. Plant Signal Behav. 2009; 4(10): 1002–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen H, Li Z, Xiong L: A plant microRNA regulates the adaptation of roots to drought stress. FEBS Lett. 2012; 586(12): 1742–7. PubMed Abstract | Publisher Full Text\n\nJaffe MJ, Takahashi H, Biro RL: A pea mutant for the study of hydrotropism in roots. Science. 1985; 230(4724): 445–7. PubMed Abstract | Publisher Full Text\n\nTakahashi N, Yamazaki Y, Kobayashi A, et al.: Hydrotropism interacts with gravitropism by degrading amyloplasts in seedling roots of Arabidopsis and radish. Plant Physiol. 2003; 132(2): 805–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlilou I, Xu J, Wildwater M, et al.: The PIN auxin efflux facilitator network controls growth and patterning in Arabidopsis roots. Nature. 2005; 433(7021): 39–44. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSengupta D, Reddy AR: Water deficit as a regulatory switch for legume root responses. Plant Signal Behav. 2011; 6(6): 914–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nDixit S, Kumar Biswal A, Min A, et al.: Action of multiple intra-QTL genes concerted around a co-localized transcription factor underpins a large effect QTL. Sci Rep. 2015; 5: 15183. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenry A, Cal AJ, Batoto TC, et al.: Root attributes affecting water uptake of rice (Oryza sativa) under drought. J Exp Bot. 2012; 63(13): 4751–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang B, Fry JD: Root Anatomical, Physiological, and Morphological Responses to Drought Stress for Tall Fescue Cultivars. Crop Sci. 1998; 38(4): 1017–1022. Publisher Full Text\n\nSteudle E: Water uptake by roots: effects of water deficit. J Exp Bot. 2000; 51(350): 1531–42. PubMed Abstract | Publisher Full Text\n\nChaves MM, Oliveira MM: Mechanisms underlying plant resilience to water deficits: prospects for water-saving agriculture. J Exp Bot. 2004; 55(407): 2365–84. PubMed Abstract | Publisher Full Text\n\nMeyer RF, Boyer JS: Sensitivity of cell division and cell elongation to low water potentials in soybean hypocotyls. Planta. 1972; 108(1): 77–87. PubMed Abstract | Publisher Full Text\n\nSerraj R, Sinclair TR: Osmolyte accumulation: can it really help increase crop yield under drought conditions? Plant Cell Environ. 2002; 25(2): 333–41. PubMed Abstract | Publisher Full Text\n\nOosterhuis DM, Wullschleger SD: Osmotic Adjustment in Cotton (Gossypium hirsutum L.) Leaves and Roots in Response to Water Stress. Plant Physiol. 1987; 84(4): 1154–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSauter A, Davies WJ, Hartung W: The long-distance abscisic acid signal in the droughted plant: the fate of the hormone on its way from root to shoot. J Exp Bot. 2001; 52(363): 1991–7. PubMed Abstract | Publisher Full Text\n\nAshraf M, Foolad MR: Roles of glycine betaine and proline in improving plant abiotic stress resistance. Environ Exp Bot. 2007; 59(2): 206–16. Publisher Full Text\n\nYu LH, Wu SJ, Peng YS, et al.: Arabidopsis EDT1/HDG11 improves drought and salt tolerance in cotton and poplar and increases cotton yield in the field. Plant Biotechnol J. 2016; 14(1): 72–84. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nHe X, Chen Z, Wang J, et al.: A sucrose: Fructan-6-fructosyltransferase (6-SFT) gene from Psathyrostachys huashanica confers abiotic stress tolerance in tobacco. Gene. 2015; 570(2): 239–47. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGe L, Chao D, Shi M, et al.: Overexpression of the trehalose-6-phosphate phosphatase gene OsTPP1 confers stress tolerance in rice and results in the activation of stress responsive genes. Planta. 2008; 228(1): 191–201. PubMed Abstract | Publisher Full Text\n\nAhn S, Anderson JA, Sorrells ME, et al.: Homoeologous relationships of rice, wheat and maize chromosomes. Mol Gen Genet. 1993; 241(5–6): 483–90. PubMed Abstract\n\nHirota O, Oka M, Takeda T: Sink activity estimation by sink size and dry matter increase during the ripening stage of barley (Hordeum vulgare) and rice (Oryza sativa). Ann Bot. 1990; 65(4): 349–353. Reference Source\n\nGladun IV, Karpov EA: Production and partitioning of assimilates between the panicle and vegetative organs of rice after flowering. Russ J Plant Physiol. 1993; 40(5): 629-633. Reference Source\n\nLi X, Lawas LMF, Malo R, et al.: Metabolic and transcriptomic signatures of rice floral organs reveal sugar starvation as a factor in reproductive failure under heat and drought stress. Plant Cell Environ. 2015; 38(10): 2171–92. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nNguyen GN, Hailstones DL, Wilkes M, et al.: DROUGHT STRESS: Role of Carbohydrate Metabolism in Drought-Induced Male Sterility in Rice Anthers. J Agron Crop Sci. 2010; 196(5): 346–57. Publisher Full Text\n\nSheoran IS, Saini HS: Drought-induced male sterility in rice: Changes in carbohydrate levels and enzyme activities associated with the inhibition of starch accumulation in pollen. Sexual Plant Reprod. 1996; 9(3): 161–9. Publisher Full Text\n\nJin Y, Yang H, Wei Z, et al.: Rice male development under drought stress: phenotypic changes and stage-dependent transcriptomic reprogramming. Mol Plant. 2013; 6(5): 1630–45. PubMed Abstract | Publisher Full Text\n\nTuberosa R: Phenotyping for drought tolerance of crops in the genomics era. Front Physiol. 2012; 3: 347. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNguyen TT, Klueva N, Chamareck V, et al.: Saturation mapping of QTL regions and identification of putative candidate genes for drought tolerance in rice. Mol Genet Genomics. 2004; 272(1): 35–46. PubMed Abstract | Publisher Full Text\n\nDu H, Wu N, Fu J, et al.: A GH3 family member, OsGH3-2, modulates auxin and abscisic acid levels and differentially affects drought and cold tolerance in rice. J Exp Bot. 2012; 63(18): 6467–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOh SJ, Kim YS, Kwon CW, et al.: Overexpression of the Transcription Factor AP37 in Rice Improves Grain Yield under Drought Conditions. Plant Physiol. 2009; 150(3): 1368–79. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHe Y, Hao Q, Li W, et al.: Identification and Characterization of ABA Receptors in Oryza sativa. PLoS One. 2014; 9(4): e95246. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGuo C, Luo C, Guo L, et al.: OsSIDP366, a DUF1644 gene, positively regulates responses to drought and salt stresses in rice. J Integr Plant Biol. 2016; 58(5): 492–502. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nXiong H, Li J, Liu P, et al.: Overexpression of OsMYB48-1, a Novel MYB-Related Transcription Factor, Enhances Drought and Salinity Tolerance in Rice. PLoS One. 2014; 9(3): e92913. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChae M, Lee JS, Nam MH, et al.: A rice dehydration-inducible SNF1-related protein kinase 2 phosphorylates an abscisic acid responsive element-binding factor and associates with ABA signaling. Plant Mol Biol. 2007; 63(2): 151–69. PubMed Abstract | Publisher Full Text\n\nZhang S, Haider I, Kohlen W, et al.: Function of the HD-Zip I gene Oshox22 in ABA-mediated drought and salt tolerances in rice. Plant Mol Biol. 2012; 80(6): 571–85. PubMed Abstract | Publisher Full Text\n\nHu H, Dai M, Yao J, et al.: Overexpressing a NAM, ATAF, and CUC (NAC) transcription factor enhances drought resistance and salt tolerance in rice. Proc Natl Acad Sci U S A. 2006; 103(35): 12987–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu H, You J, Fang Y, et al.: Characterization of transcription factor gene SNAC2 conferring cold and salt tolerance in rice. Plant Mol Biol. 2008; 67(1–2): 169–81. PubMed Abstract | Publisher Full Text\n\nYou J, Hu H, Xiong L: An ornithine δ-aminotransferase gene OsOAT confers drought and oxidative stress tolerance in rice. Plant Sci. 2012; 197: 59–69. PubMed Abstract | Publisher Full Text\n\nXiang Y, Tang N, Du H, et al.: Characterization of OsbZIP23 as a key player of the basic leucine zipper transcription factor family for conferring abscisic acid sensitivity and salinity and drought tolerance in rice. Plant Physiol. 2008; 148(4): 1938–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJisha V, Dampanaboina L, Vadassery J, et al.: Overexpression of an AP2/ERF Type Transcription Factor OsEREBP1 Confers Biotic and Abiotic Stress Tolerance in Rice. PLoS One. 2015; 10(6): e0127831. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLiu C, Mao B, Ou S, et al.: OsbZIP71, a bZIP transcription factor, confers salinity and drought tolerance in rice. Plant Mol Biol. 2014; 84(1–2): 19–36. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTang N, Zhang H, Li X, et al.: Constitutive activation of transcription factor OsbZIP46 improves drought tolerance in rice. Plant Physiol. 2012; 158(4): 1755–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZou M, Guan Y, Ren H, et al.: A bZIP transcription factor, OsABI5, is involved in rice fertility and stress tolerance. Plant Mol Biol. 2008; 66(6): 675–83. PubMed Abstract | Publisher Full Text\n\nYoshida T, Fujita Y, Sayama H, et al.: AREB1, AREB2, and ABF3 are master transcription factors that cooperatively regulate ABRE-dependent ABA signaling involved in drought stress tolerance and require ABA for full activation. Plant J. 2010; 61(4): 672–85. PubMed Abstract | Publisher Full Text\n\nYou J, Zong W, Li X, et al.: The SNAC1-targeted gene OsSRO1c modulates stomatal closure and oxidative stress tolerance by regulating hydrogen peroxide in rice. J Exp Bot. 2013; 64(2): 569–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCai S, Jiang G, Ye N, et al.: A key ABA catabolic gene, OsABA8ox3, is involved in drought stress resistance in rice. PLoS One. 2015; 10(2): e0116646. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation"
}
|
[
{
"id": "14735",
"date": "30 Jun 2016",
"name": "Krishna Jagadish",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14736",
"date": "30 Jun 2016",
"name": "Zhulong Chan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14737",
"date": "30 Jun 2016",
"name": "Arvind Kumar",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1554
|
https://f1000research.com/articles/5-1553/v1
|
30 Jun 16
|
{
"type": "Review",
"title": "Recent advances in imaging subcellular processes",
"authors": [
"Kenneth A. Myers",
"Christopher Janetopoulos",
"Kenneth A. Myers"
],
"abstract": "Cell biology came about with the ability to first visualize cells. As microscopy techniques advanced, the early microscopists became the first cell biologists to observe the inner workings and subcellular structures that control life. This ability to see organelles within a cell provided scientists with the first understanding of how cells function. The visualization of the dynamic architecture of subcellular structures now often drives questions as researchers seek to understand the intricacies of the cell. With the advent of fluorescent labeling techniques, better and new optical techniques, and more sensitive and faster cameras, a whole array of questions can now be asked. There has been an explosion of new light microscopic techniques, and the race is on to build better and more powerful imaging systems so that we can further our understanding of the spatial and temporal mechanisms controlling molecular cell biology.",
"keywords": [
"imaging subcellular processes",
"microscope",
"cell biology",
"Stimulated emission-depletion microscopy",
"STED",
"microscopy techniques"
],
"content": "Introduction\n\nOptical microscopy is a diffraction-limited imaging methodology. As such, the use of light to generate a microscopic image is inherently bound by the wavelength(s) of light used to generate a diffraction pattern and the aperture of the objective lens oriented to collect the diffracted light source. These limitations were first described mathematically by Ernst Karl Abbe as a lateral resolution limit equivalent to the wavelength of light divided by two times the numerical aperture of the objective lens (d = λ/2NA)1. With this fundamental resolution limit established and verified experimentally, a century of light microscopy working at or near the limit of light diffraction has resulted in the identification of numerous cell biological phenomena2–7.\n\nFor the purposes of conducting experimental imaging of cell biological specimens, the resolution limitations of light microscopy are somewhat balanced by the advantages inherent in the light microscopic technique. For example, it is possible to exceed the light diffraction limit and thereby achieve super-resolution images by using techniques such as atomic force microscopy, scanning tunnel microscopy8, shear force microscopy, or other common types of scanning probe microscopy (SPM)9. In addition, tried and true techniques such as electron microscopy (EM) generate extremely high-resolution images (sub-nanometer resolution), yet each of these imaging modalities manages to overcome the diffraction limit and achieve sub-diffraction limit resolution images by avoiding the use of visible light waves altogether. Additionally, the use of SPM and EM imaging methodologies for biological investigations often involves complicated and expensive sample preparation procedures. More importantly, because SPM and EM methods require physical “scanning” of the biological sample or require a fixed (non-living) sample, these super-resolution techniques have extremely low temporal resolution and do not allow imaging of the coordinated and dynamic movements of organelles, proteins, or other biological molecules located within subcellular processes. Thus, as experimental microscopists continue to use diffraction-limited light microscopy to piece together the finer details of the microscale architecture and dynamics of the cell, the need to break the light diffraction limit has become a focal point for biologists interested in defining the nanoscale organization of the cell with coordinately high spatial and temporal resolution.\n\nThis article reviews recent advances in super-resolution microscopy techniques as applied to microscopic imaging of subcellular processes. Ten to fifteen years ago, the concepts that underlie many of these super-resolution techniques were in place and were actively being tested, often with moderate to significant effect in increasing the resolution of biological samples and breaking the diffraction limit described by Abbe nearly 150 years before. In more recent years, super-resolution microscopy has continued down this path, heightening our ability to discern the biological details of primarily fluorescent molecules to single-nanometer resolution. Here, we discuss a variety of super-resolution imaging techniques first by describing the theoretical concepts that underlie the imaging methodology and then by providing examples of how some of these technologies have been applied to study subcellular biological processes. What is clear from the literature is that there has been an explosion of new imaging methods. As a result, we were forced to focus on a subset of the available super-resolution imaging approaches, and we apologize to the authors of the many optical techniques not discussed in this review.\n\n\nStimulated emission-depletion microscopy\n\nThe super-resolution microscopic technique termed stimulated emission-depletion (STED) was first described as a far-field imaging technique capable of producing 3D images with 35 nm resolution10. STED microscopy operates by using two laser beams to illuminate the specimen. The first laser beam is an excitation laser pulse that is immediately followed (<1 ps) by a second red-shifted laser that illuminates the sample with a donut-shaped pattern. The STED approach is based on the theory that the full width at half maximum (FWHM) of the point spread function (PSF) can be reduced by the second laser beam that effectively takes a subpopulation of excited fluorescent molecules (from the first beam) and returns them to the ground state before they can spontaneously fluoresce. To achieve this, the STED laser intensity must be powerful enough that the stimulated emission can outcompete the fluorescence emission. Because the STED beam generates a non-linear depletion, the center of the beam has zero intensity while the periphery of the beam has high intensity (the donut-shaped pattern;11; Figure 1). Thus, the excited fluorescent molecules in the periphery of the beam are turned “off” while those in the center remain “on”. The outcome is a narrowed PSF with an achieved FWHM approaching a range of 20 to 60 nm. The STED technique has been used to generate super-resolution images revealing that synaptotagmin remains clustered after synaptic vesicle exocytosis (Figure 1).\n\n(a) Principles of operation. While the blue excitation (EXC) beam is focused to a diffraction-limited excitation spot, shown in the adjacent panel in blue, the orange STED beam is able to de-excite molecules. The STED beam is phase-modulated to form the focal dounut shown in the top right panel. Superimposition of the two focal spots confines the area in which fluorescence is possible to the dounut center, yielding the effective fluorescent spot of subdiffraction size shown in green in the lower panel. All spots represent measured data and are drawn to scale. The profile of the green effective fluorescent spot has a full width at half maximum of 66 nm as well as a sharp peak. The green spot shows an 11-fold reduction in focal area beyond the excitation diffraction value (compare with blue spot). (b) Mechanism of synaptic labeling. Synaptic vesicles exocytose, allowing their lumenal synaptotagmin domains to bind anti-synaptotagmin antibodies. These antibodies are internalized upon endocytosis. (c) Typical image of a neuron labeled with an anti-synaptotagmin antibody, fixed, permeabilized, and visualized by using Atto532-labeled secondary antibodies. Fluorescent puncta represent labeled synaptic nerve terminals. Scale bar = 10 mm. (d) Comparison of confocal (left) and STED (right) counterpart images of a labeled preparation reveals a marked increase in resolution by STED. Scale bar = 500 nm. PSF, point spread function. Taken from 11.\n\nIn addition to the ability of STED to capture spatial and temporal super-resolution images, multicolor STED can investigate several different molecules simultaneously12. Live-cell investigations using STED microscopy have been performed to investigate the dynamics of the syntaxin-1A protein with presynaptic vesicles13 to define the super-resolution map of the ciliary transition zone14 and to measure vesicular motility in living Drosophila15. Recent studies have used STED microscopy to investigate the subcellular dynamics of the Bcl-2 family of mitochondrial proteins, Bax and Bak. This study found that when cells enter apoptosis, activated Bax molecules form large and compact clusters that assemble with Bak into ring-like structures in the mitochondrial outer membrane. These assemblies then contribute to the formation of large pores and suggest a mechanism for outer membrane permeabilization by Bax16.\n\nAchieving optimal super-resolution microscopic images depends largely on choosing the best tool(s) for the job. As with each super-resolution imaging approach, advantages of the STED technique are coupled with disadvantages. For biological imaging applications, the implementation of STED microscopy requires the use of STED probes that are specifically targeted to the protein of interest. Additionally, the fluorescent properties of the probes must be carefully evaluated because the resolution of STED relies on the characteristic saturation intensity and photostability of the fluorophore17,18. As a result, a relatively small number of fluorescent proteins and organic dyes are useful for STED imaging. Commonly used approaches for fluorescent labeling of proteins in live-cell applications, such as the green fluorescent protein, can be used in some cases for STED imaging; however, the brightness and quantum yield are often limiting, and there are few STED options for genetically encodable fluorescent labels in the far-red region, thus reducing the number of colors that can be imaged within the same specimen.\n\n\nStructured illumination microscopy\n\nStructured illumination microscopy (SIM) uses sample illumination with spatially structured excitation light of a known orientation. This approach makes normally inaccessible high-resolution information visible in the observed image. A series of images is reciprocally processed by using a Fourier transformation of the structured excitation image in order to generate a reconstruction with improved resolution. Preliminary experiments have confirmed the validity of the physical principle and the capability of this imaging methodology by using both test objects and complex biological structures to demonstrate superior effective resolution compared with conventional and confocal microscopes (Figure 2)19–21. Performed in its linear form, SIM allows a doubling of the lateral resolution of a biological sample21. Non-linear forms of SIM, for example, saturated pattern excitation microscopy and saturated SIM, couple linear SIM with ground-state depletion techniques resulting in lateral resolution of objects as small as 50 nm22.\n\n(a) Widefield (top) and SI images (bottom) of 40 nm beads. Bar = 1 μm. (b) Plot profiles from the region marked by red bars in (a) show that two beads approximately 200 nm apart can be resolved only following SI reconstruction. Fitting of the curve with two Gaussians gives full widths at half maximum (FWHMs) (red bars on the graph) of 99 and 105 nm for each bead, respectively. (c) Confocal (left), widefield (middle), and SI image (right) of human primary natural killer cells activated on a surface coated with anti-NKG2D monoclonal antibody. Bars = 5 μm. (d) Regions at the center of the synapses in (c) are enlarged to demonstrate the increased level of detail in cortical F-actin structure when SI microscopy is used. Bars = 1 μm. (e) Plot profiles to directly compare the widefield and SI reconstructed image for the region indicated by the red bar shown in (d). Estimation of the FWHM (red bar on the graph) from the Gaussian fit of the SI data gives a resolution of approximately 115 nm. a.u., arbitrary units. Taken from 19.\n\nWhen the concept of structured illumination was first applied to imaging of biological samples21, the approach used two superimposed line patterns of known geometry, which mathematically results in the multiplication of the line patterns with the fluorescently labeled sample structure, with the product containing moiré fringes. The method of imaging subcellular processes can be combined with Fourier transforms by overlapping two different spatial frequencies from multiple directions to obtain finer spatial frequencies emitted by the specimen. By using structured illumination to acquire a sequence of images, each with different orientation and phase of the pattern, information is recovered from twice the area of what is observable using confocal microscopy, and a resulting doubling of the resolution, including lateral resolutions in the range of 100 nm and axial resolutions approaching 300 nm.\n\nRecent advances using structured illumination technology have combined non-fluorescent Raman imaging with SIM technology to increase the spatial resolution of both biological and inorganic chemical mapping23. Additionally, investigations of the subcellular structures termed focal adherens junctions (FAJs) have used SIM to identify that tension-sensitive focal adhesion protein members display distinct localization patterns within FAJs that may function as a force-sensitive module capable of regulating specific aspects of junction dynamics24.\n\nOther experimental investigations have incorporated 3D two-color SIM datasets of living cells to visualize super-resolution time-lapse dynamics of mitochondria, vesicles, and the actin cytoskeleton on the order of 8.5 seconds per two-color image (4 seconds for single color)25. As with all existing super-resolution microscopic techniques, efforts toward enhancing SIM have focused on both increased spatial resolution and enhanced temporal resolution needed to visualize dynamics of subcellular components. Examples of SIM-associated advances in temporal resolution include two-photon point-based scanning methods, which have the ability to acquire super-resolution images approximately one image per second26,27, in addition to the acquisition of 3D super-resolution images in thick, semi-transparent biological specimens27,28. Yet another SIM modality, termed “instant SIM” (iSIM), has achieved the most temporally resolved super-resolution imaging. The iSIM approach is optimized for live-cell imaging of 2D or 3D biological samples and is capable of achieving 145 nm lateral and 320 nm axial resolution at frame rates faster than 100 frames per second29–31.\n\n\nSingle-molecule imaging modalities\n\nSome of the first applications for single-molecule detection combined optical force traps with single-molecule fluorescence techniques to characterize nucleotide dynamics and force production of myosin motors32–34. Subsequently, a theoretical concept underlying single-molecule imaging was demonstrated in 2003 and was termed fluorescence imaging with one-nanometer accuracy (FIONA)35. This methodology was used to perform single-molecule measurements of the myosin V molecular motor protein and showed in exquisite detail that myosin V walks along actin filaments in a hand-over-hand mechanism. Perhaps of critical importance to the future of super-resolution microscopy, the FIONA technique showed how fluorescent molecules could be localized with sub-diffraction precision by curve-fitting fluorescence emission data (the PSF) to determine the mean value of the emission distribution and the corresponding standard deviation of the distribution. The PSF is defined by four variables that also define the diffraction limit of resolving two microscopic points or molecules: (1) the number of collected photons (highly influenced by the numerical aperture of the objective lens), (2) the pixel size of the detector, (3) the standard deviation of the background noise plus detector noise, and (4) the standard deviation of the signal (this determines the precision of the PSF, also known as the FWHM).\n\nLimitations of FIONA have since been overcome by modification of this imaging approach. For example, FIONA can measure a single molecular domain over time but cannot resolve multiple fluorescence emitters in close proximity. The commonality of multiple emitters led to further developments in the theoretical technique in order to eliminate or at least to reduce this obstacle. Photo-bleaching of a subset of fluorescent emitters was one methodology capable of reducing the multiple-emitter problem. Soon, numerous adaptations of FIONA were employed to answer important cell biological questions that required sub-diffraction resolution of single molecules, including nanometer-localized multiple single-molecule fluorescence microscopy36, single-molecule high-resolution imaging with photo-bleaching37, single-molecule high-resolution co-localization microscopy38, stochastic optical reconstruction microscopy (STORM)39, and photo-activated localization microscopy (PALM)40. In both of these single-molecule imaging modalities, the addition of turning off the fluorescent probe via photo-switching or photo-bleaching was the key underlying principle that made these techniques capable of distinguishing multiple fluorescent emitters within a biological sample.\n\nPALM imaging took the FIONA concept of curve-fitting a PSF to generate a super-resolution image and combined it with photo-activatable or photo-switchable (PA) fluorescent molecules40. This approach is ideal for generating high signal-to-noise ratios because the sample is essentially “dark” until the PA wavelength of light is applied to the sample. Additionally, the PA approach permits controlled levels of photo-activation (by modifying activation laser light intensity or exposure times) of whole cells or within sub-cellular processes (using an imaging system capable of illuminating user-defined regions of interest).\n\nPALM imaging uses a repetitive process in which a subset of the PA fluorophores within a biological sample is first excited, an image is acquired, and then the excited fluorophores are turned off by either photo-switching or photo-bleaching the fluorescent probe(s). This approach allows a small number of fluorophores to emit photons with each round of imaging. After repetition of this process potentially hundreds to thousands of times on a single sample, the distribution and localization intensity of each acquired PSF can be assigned computationally to generate the super-resolution image40. The ultimate resolution achievable in PALM is dependent on the localization precision (how well the center of each PSF can be determined) and the density of available PA molecules. Thus, the four contributing PSF variables (described above) apply to PALM imaging, as well as the density of the PA molecules. Both 2D and 3D PALM techniques are complemented by the use of total internal reflection fluorescence (TIRF) microscopy41,42. Although TIRF is not theoretically required to perform PALM imaging, the advantage of TIRF is the use of an evanescent wave of illumination that generates a relatively narrow depth of field (~200 nm) for imaging. Thus, TIRF microscopy inherently limits the amount of out-of-focus light that reaches the objective lens and thereby increases the resolution of the resulting image. The high signal-to-noise ratio gained by TIRF microscopy is excellent for imaging focal adhesions or membrane-associated proteins that are within the 200 nm evanescent illumination plane, but it also highlights the limitations of the TIRF approach to a narrow depth of imaging field.\n\nAt the time it was first developed, one inherent limitation of the PALM technique was that it required the biological sample to be fixed (non-living) and thereby removed the possibility of live-imaging the dynamics of the system. Adaptations of the original PALM technique have since allowed live-cell investigations to obtain measurements of nanoscale dynamics of adhesion complexes43, T-cell signaling44, and a number of membrane-bound protein investigations45–47. Still, the major limitation of PALM imaging is low temporal resolution because the need for spatial distinction of photo-activated molecules, which by its nature results in low quantum yield, requires increased acquisition time to overcome this obstacle. Therefore, live-cell PALM imaging provides substantially increased spatial resolution but suffers from low temporal resolution and therefore has not been used to image temporally dynamic subcellular events. However, new innovations of PALM have made breakthroughs that significantly reduce the time necessary to acquire PALM images and provide hope for continued enhancements in temporal resolution using this imaging modality48.\n\nPALM imaging of subcellular processes has been used to conduct 2D analysis of focal adhesion complexes43 and, using a PALM modification termed “interferometric” PALM (iPALM), to investigate the 3D structure of focal adhesion complexes49,50. Two-dimensional studies of focal complexes using a single- or dual-color PALM approach have shown the stratified organization of the focal adhesion molecules vinculin and alpha-actinin in super resolution. These results were compared with images of the same fluorescent molecules within the same adhesion structures to highlight the enhanced resolution of the PALM technique.\n\nTo apply the PALM technique to investigations of the 3D organization of subcellular structures, investigators developed a modification termed iPALM, which is the combination of PALM with single-photon, simultaneous multiphase interferometry that provides sub-20 nm 3D protein localization with optimal molecular specificity49. Using the well-defined principle of self-interference of a single-photon source, iPALM orients a biological sample containing PA-labeled molecules at the focal plane of two opposed objective lenses in order to allow for the precise determination of each molecule’s axial position (Figure 3). Upon illumination, a fluorescent photon simultaneously enters both the upper and lower objectives and the difference in path lengths of the upper and lower beams directly depends on the axial position of the source. Additionally, self-interference in the three-way beam splitter results in an axial-dependent modulation of the relative intensities of the three output beams. This concept enabled the 3D position of the source molecule to be determined from the relative amplitudes of the source images from the three cameras (Figure 3). This technology was most prominently applied in a study that mapped the nanoscale organization of focal adhesion proteins (Figure 4)50.\n\n(a, b) Schematic of the single-photon multiphase fluorescence interferometer. A point source with z-position δ emits a single photon both upwards and downwards. These two beams (color-coded as red and green in (b) interfere in a special three-way beam splitter. (c) The self-interfered photon propagates to the three color-coded charge-coupled device (CCD) cameras with amplitudes that oscillate 120° out of phase as indicated. a.u., arbitrary units; BP, band pass; NA, numerical aperture. Taken from 49.\n\nTop view and side view “interferometric” photo-activated localization microscopy (iPALM) images of focal adhesions (white boxes, top-view panels) and corresponding z histograms and fits. (a, b) FAK. (c, d) Paxillin. (e, f) Vinculin. (g, h) Zyxin. (i, j) VASP. (k, l) Alpha-actinin. The vertical distribution of α-actinin is non-Gaussian, so the focal adhesion peak fit is not shown. Paxillin and α-actinin shown are C-terminal photo-activatable fluorescent protein (PA-FP)-tagged. Colors: vertical (z) coordinate relative to the substrate (z = 0 nm, red). (d, bottom panel). Schematic model of focal adhesion molecular architecture depicts experimentally determined protein positions. Note that the model does not depict protein stoichiometry. Scale bars = 5 μm (a, c, e, g, i, k) and 500 nm (b, d, f, h, j, l). ECM, extracellular matrix. Taken from 50.\n\nIn the same way that PALM imaging uses stochastic activation of PA fluorescent probes, the fundamental principle behind STORM is that the activated state of a photo-switchable molecule must lead to the consecutive emission of sufficient photons to enable precise localization before it enters a dark state or becomes deactivated by photo-bleaching. Additionally, the sparsely activated fluorescent molecules must be separated by a distance that exceeds the Abbe diffraction limit (approximately 250 nm) to enable the parallel recording of many individual emitters, each having a distinct set of coordinates in the lateral image plane.\n\nBiological applications of the STORM imaging methodology (Figure 5) have revealed that axonal actin is organized as ring-like structures spaced every 180 to 190 nm and that this periodic arrangement of actin is linked to the organization of spectrin and ankyrin proteins within a region of the neuron termed the axon initial segment (AIS)51. The physiological significance of this organization in neurons is proposed to somehow operate as a transport “scaffold” or “filter” that functions to specifically distinguish the AIS from other regions of the neuron to maintain the identity of the axon52–54. The fine molecular details of the AIS scaffold were recently identified by using multicolor 2D- and 3D-STORM of endogenous epitopes to demonstrate that the periodic organization is composed of longitudinal head-to-head βIV-spectrin subunits that work to connect actin-rich bands along the AIS55. Furthermore, this study used STORM imaging to identify that ankyrin G scaffold protein is associated with the AIS membrane, where it extends approximately 35 nm into the membrane-adjacent cytoplasm in order to possibly facilitate an interaction with peripheral microtubules and thereby regulate vesicular entry into the axon (Figure 5)55. These data suggest that the organization of AIS components is necessary for defining axon integrity by establishing a subcellular gateway to the axon.\n\n(a) The schematic shows that head-to-head βIV-spectrins connect the actin rings to form a periodic sub-membrane complex while ankyrin G C-terminal tails extend 32 nm below the sub-membrane complex to allow possible interactions with the microtubule cytoskeleton that are important for establishing and maintaining axonal identity. (b, c) STORM images of neurons treated with vehicle (dimethyl sulfoxide 0.1%; 1 hour; b, c) and then fixed and labeled for actin (b) or βIV-spectrin specific domain (SD) (c). Scale bar = 2 μm. Taken and modified from 55.\n\nSTORM imaging has also been used to discern fine details of the microtubule-organizing center, including the pericentriolar material (PCM), which has remained elusive in its composition for decades56. This study combined SIM and STORM imaging to attempt to identify the organization of the PCM in Drosophila. STORM imaging revealed that the Pericentrin-like protein (Plp) is distributed in distinct molecular clusters around the centriole and forms molecular fibrils extending into the PCM matrix, comprising a scaffold similar to spokes of a wheel. The combined SIM/STORM study demonstrated that the PCM is composed of two distinct structural layers that provide separate functionality. Moreover, the investigators discovered that Plp plays an important role in PCM organization by forming a molecular scaffold for the PCM matrix56. The combination of these two super-resolution methodologies was necessary because of the limitations of STORM sample preparation57. The ability to overcome these limitations or to invent new feasible methods of sample preparation will likely determine the utility of STORM imaging of nanoscopic cellular structures.\n\n\nImaging live specimens\n\nDuring traditional epifluorescence, confocal, and many of the super-resolution microscopy techniques, the specimen is illuminated with an extensive beam that travels through the whole sample. The problem with this light path is that the two cones of light excite the out-of-focus components of the sample. While techniques such as spinning disk confocal microscopy that allow for full-field confocal imaging have helped with live-specimen imaging, any excess light as found in these techniques is damaging to the sample, releasing free oxygen radicals, bleaching the fluorescent probes, and even heating the specimen. This often will cause artifacts such as cell retraction or shrinkage, and may result in cell death. Many of the techniques described above are still not very useful for live-cell imaging, often requiring that the cells be fixed. Better techniques are being developed that provide the ability to image many frames over short durations for the examination of dynamic cellular events. These four-dimensional (4D) techniques can also be valuable for obtaining more subtle changes occurring over many hours or even days.\n\n\nLight sheet fluorescence microscopy\n\nTo help address this problem of specimen photo-damage, several techniques have been developed in which only the imaging plane of focus is illuminated with a “sheet” of light. Planar illumination was first proposed by the 1925 chemistry Nobel Prize winner Richard Adolf Zsigmondy along with German physicist Heinrich Siedentopf in 190258. The technique was revitalized in the late 1990s when microscopists realized its potential for experimental use as new fixed and live-cell fluorescent probes came on line for imaging subcellular organelles and molecules. In 2004, Huisken et al. published an article demonstrating the viability of selective plane illumination microscopy (SPIM), which is synonymous with light sheet fluorescence microscopy (LSFM)59. These systems typically have an imaging objective and an orthogonal cylindrical lens or excitation objective that excites the specimen with a thin sheet of light. Because optical sectioning is achieved by the excitation light sheet, an advantage of SPIM is that the entire focal plane is imaged simultaneously (Figure 6a). SPIM was found to be useful for imaging large, multicellular specimens, but the benefits of using SPIM for single-cell microscopy are limited. This is because of the inherent nature of the illumination and the divergence of the beam. The illumination can become thinner in only a relatively small region, typically in the z direction. The thinner the sheet is, the narrower it becomes in x or y, and the faster it diverges (fattens) elsewhere in the sample. This does not allow the sheet to get much flatter than a cell before it becomes a focused point of light. Nevertheless, this configuration permits fast imaging of an entire plane that can be imaged with sensitive complementary metal-oxide-semiconductor or EM charge-coupled device cameras. Whole volume collection can be acquired either by scanning the sample through a stationary objective or by moving the objective synchronously to scan a stationary sample. This has been useful for imaging large specimens.\n\n(a) The traditional approach, in which a Gaussian beam is swept across a plane to create the light sheet. (b) A Bessel beam of comparable length produces a swept sheet with a much narrower core but flanked by sidebands arising from concentric side lobes of the beam. (c, d) Bound optical lattices create periodic patterns of high-modulation depth across the plane, greatly reducing the peak intensity and the photo-toxicity in live-cell imaging. The square lattice in (c) optimizes the confinement of the excitation to the central plane, and the hexagonal lattice in (d) optimizes the axial resolution as defined by the overall point spread function (PSF) of the microscope. The columns in (a to d) show the intensity pattern at the rear pupil plane of the excitation objective; the cross-sectional intensity of the pattern in the xz plane at the focus of the excitation objective (scale bar = 1.0 μm); the cross-sectional intensity of the light sheet created by dithering the focal pattern along the x axis (scale bar = 1.0 μm); and the xz cross-section of the overall PSF of the microscope (scale bar = 200 nm). Taken from 69. Abbreviations:AU; arbitrary units.\n\n\nBessel beam microscopy\n\nThe development of Bessel beam microscopy has provided a solution for imaging smaller features and cells by creating a beam that can illuminate the sample in a planar manner60,61. Bessel beams are a class of non-diffracting beams that will maintain a relatively tight and unabberated focus over a prolonged distance of approximately 50 μm62,63. The Bessel beam enters the lens only on the periphery of the lens, creating a funnel of light with a flat focus point (Figure 6b). Despite the flat focus of the Bessel beam, imaging studies using this technique revealed problems with out-of-focus excitation. To overcome this obstacle, two-photon excitation has been introduced to the Bessel beam system63–66. Two-photon excitation diminishes by the square of the excitation intensity, and the outer-lying lobes of the Bessel beam become less relevant67. This approach comes with all the typical limitations of two-photon microscopy, including the cost, but does allow the deeper imaging into tissue specimens.\n\nAdditional functionalities were added to the Bessel beam approach by creating a Bessel beam super-resolution SIM, which adds the principles of 3D super-resolution SIM (3D SR-SIM)65,68. The Bessel beam SR-SIM has an axial resolution that is approximately two times that of confocal techniques and roughly similar to that of widefield SR-SIM.\n\n\nLattice light sheet microscopy\n\nMore recently, the Betzig lab has engineered a Bessel beam microscope with multiple beams that allow the creation of a lattice light sheet microscope69. Optical lattices are periodic interference patterns in two or three dimensions created by the coherent superposition of a finite number of plane waves travelling in certain well-defined directions. An ideal 2D lattice is non-diffracting in the sense that it propagates indefinitely in a direction y without changing its cross-sectional profile, which extends infinitely in x and z. This is accomplished by confining the illumination at the rear pupil plane of the excitation objective to points on an infinitesimally thin ring. This allows one to image a specimen with the same total flux as measured with individual spots in confocal microscopy, but that flux is dispersed among the multiple beams and the cell is exposed to reduced amounts of excitation illumination (Figure 6d)69. This lattice light sheet microscopy (LLSM) had clear advantages over the single Bessel beam microscope, including an approximately 75% reduction in the total radiation delivered and the ability to image approximately three times faster for patterns of a similar period. This approach also allows imaging with finer patterns that require fewer steps per plane, resulting in a 3D volume acquisition that is several times faster than with the Bessel approach69. Chen et al.69 (2014) demonstrated the advantages of the LLSM by imaging 20 different specimens spanning four orders of magnitude in space and time, including the binding kinetics of individual transcription factors, 3D super-resolution PALM of nuclear lamins, dynamic organelle rearrangements and 3D tracking of microtubule plus ends during mitosis, neutrophil motility in a collagen mesh, and subcellular protein localization and dynamics during embryogenesis in Caenorhabditis elegans and Drosophila melanogaster. As an example of the minimal amount of phototoxicity elicited on a specimen, a single ciliated Tetrahymena thermophila cell was imaged continuously for over 7 minutes, and over 200,000 frames were acquired. In 2D mode, 18,000 frames were acquired at 3-millisecond intervals highlighting the beating of cilia and the movement of subcellular organelles. In addition, full 3D volumes of the microtubule cytoskeleton in a Dictyostelium discodeum cell were acquired for 1,430 frames, and volumes were taken every 2 seconds, and there was no noticeable photobleaching or phototoxicity69.\n\n\nInverted selective plane illumination microscope\n\nOne limitation of the above LLSM approaches is the arrangement of the sample holder, which limits the type of samples you can observe and often makes interacting with the sample with micromanipulators or even perfusion difficult. An easier sample mounting approach was developed with the inverted selective plane illumination microscopy (iSPIM), which uses a standard inverted microscope platform70. For excitation, a laser passes through a mask and a mirror directs it into a 45° oriented excitation objective mounted on an anchor pillar which directs the excitation light sheet toward the specimen. The fluorescence is collected by a second 45° oriented objective. A second camera can be positioned under the inverted microscope platform objectives for focus-finding and transmitted light purposes. This type of imaging was used to collect volumes every 2 seconds for a 14-hour period of embryogenesis (collecting approximately 25,000 volumes) with no detectable phototoxicity in a C. elegans nematode. They also performed two-color imaging that allowed them to perform lineage tracing while also visualizing neurodevelopmental dynamics70. Other groups have used similar LLSM techniques for 4D imaging in zebrafish embryos71 and in the fruit fly Drosophila melanogaster72,73. The iSPIM has also been modified to capture a second specimen view by alternating excitation and detection between the two objectives. The resulting “diSPIM” provides isotropic spatial resolution (down to 330 nm) at high speed (200 images per second, 0.5 seconds for a dual-view 50-plane volume) and has been used to successfully track microtubule tips in three dimensions in living cells as well as improve the imaging of nuclear dynamics during C. elegans embryogenesis74. The Shroff group also has provided step-by-step directions for the construction of a diSPIM entirely from commercially available parts75. This design is compatible with fiber-coupled laser excitation. Such fiber-based excitation facilitates alignment, making the device compatible with a broad array of commercial laser excitation sources. Several other groups have provided instructions to build LLSM systems, making it possible for biologists to build them from commercially available microscope equipment76,77.\n\n\nSpherical-aberration-assisted extended depth of field\n\nA brand-new technique called spherical-aberration-assisted extended depth-of-field light sheet microscopy (SPED-LSM) was recently published78. This method turns spherical aberration into an advantage by combining the large volumetric field of view of an extended depth of field with the optical sectioning of light sheet microscopy. The problem with the use of spherical aberration is it reduces the contrast and resolution in both the focal plane and optical axis. In part, the authors correct for this through the use of deconvolution. The SPED-LSM technique used galvanometer scanners, which eliminated the need to physically scan the detection objective. SPED enabled scanning of thousands of volumes per second, limited only by camera acquisition rate. The authors demonstrate capabilities of SPED microscopy by performing fast subcellular resolution imaging of CLARITY79 (a technique that produces structurally intact yet optically transparent tissue) mouse brains and cellular-resolution volumetric Ca2+ imaging of entire zebrafish nervous systems. It is easy to imagine integrating a SPED system with complementary optics for optogenetics so one might simultaneously record and control neural activity across portions and maybe even the entire vertebrate nervous system. This type of microscopy will have a lot of potential as camera speeds increase.\n\n\nConcluding remarks\n\nThis is an exciting time to be a cell biologist. However, with so many new types of super-resolution microscopy techniques now available, it becomes a significant challenge to determine which microscopes are useful for your studies. One must consider what type of spatial and temporal resolution is needed to answer the questions at hand. The cell type, the type of fluorophores being used, the time between acquisitions, and the duration of the experiment all can have a significant influence on the type of microscopy that can or should be performed. With fixed specimens, one can generally illuminate the samples with much more excitation light without worrying about generating artifacts. For live-specimen microscopy, sometimes going faster and seeing smaller details provides useful information; other times it does not. One still must be careful not to oversample and damage the specimen even with some of these very gentle LSFM techniques. One thing that is quickly becoming a problem is the handling of the raw data generated with some of these new 4D techniques. Data management can quickly become a burden since acquiring 4D data with many of these advanced techniques can quickly lead to the generation of terabytes of data that need to be organized, examined, and analyzed. Once interesting data are identified, the scientific community also needs to determine better ways to make all of the data accessible. So many exciting findings are just buried away on the hard drives of research laboratory computers and never see the light of day or the right set of eyes that might recognize an important discovery. The good news is that memory storage is becoming reasonably priced, maintaining your data in “the cloud” is easier, and journals are also allowing larger sized files so that the scientific community has the ability to visualize these exciting new techniques when they are published.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAbbe E: Beiträge zur Theorie des Mikroskops und der mikroskopischen Wahrnehmung. Archiv f mikrosk Anatomie. 1873; 9(1): 413–8. Publisher Full Text\n\nGauthier A, Brandt R: Live cell imaging of cytoskeletal dynamics in neurons using fluorescence photoactivation. Biol Chem. 2010; 391(6): 639–43. PubMed Abstract | Publisher Full Text\n\nOkumoto S, Takanaga H, Frommer WB: Quantitative imaging for discovery and assembly of the metabo-regulome. New Phytol. 2008; 180(2): 271–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMichalet X, Kapanidis AN, Laurence T, et al.: The power and prospects of fluorescence microscopies and spectroscopies. Annu Rev Biophys Biomol Struct. 2003; 32: 161–82. PubMed Abstract | Publisher Full Text\n\nHaraguchi T: Live cell imaging: approaches for studying protein dynamics in living cells. Cell Struct Funct. 2002; 27(5): 333–4. PubMed Abstract | Publisher Full Text\n\nLaurent M, Johannin G, Gilbert N, et al.: Power and limits of laser scanning confocal microscopy. Biol Cell. 1994; 80(2–3): 229–40. PubMed Abstract | Publisher Full Text\n\nMüller CB, Enderlein J: Image scanning microscopy. Phys Rev Lett. 2010; 104(19): 198101. PubMed Abstract | Publisher Full Text\n\nLindblom P, Gerhardt H, Liebner S, et al.: Endothelial PDGF-B retention is required for proper investment of pericytes in the microvessel wall. Genes Dev. 2003; 17(15): 1835–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiesendanger R: Scanning-probe-based science and technology. Proc Natl Acad Sci U S A. 1997; 94(24): 12749–50. PubMed Abstract | Free Full Text\n\nHell SW, Wichmann J: Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy. Opt Lett. 1994; 19(11): 780–2. PubMed Abstract | Publisher Full Text\n\nWillig KI, Rizzoli SO, Westphal V, et al.: STED microscopy reveals that synaptotagmin remains clustered after synaptic vesicle exocytosis. Nature. 2006; 440(7086): 935–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBückers J, Wildanger D, Vicidomini G, et al.: Simultaneous multi-lifetime multi-color STED imaging for colocalization analyses. Opt Express. 2011; 19(4): 3130–43. PubMed Abstract | Publisher Full Text\n\nUllrich A, Böhme MA, Schöneberg J, et al.: Dynamical Organization of Syntaxin-1A at the Presynaptic Active Zone. PLoS Comput Biol. 2015; 11(9): e1004407. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYang B, Fang CY, Chang HC, et al.: Polarization effects in lattice-STED microscopy. Faraday Discuss. 2015; 184: 37–49. PubMed Abstract | Publisher Full Text\n\nSchneider J, Zahn J, Maglione M, et al.: Ultrafast, temporally stochastic STED nanoscopy of millisecond dynamics. Nat Methods. 2015; 12(9): 827–30. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGroße L, Wurm CA, Brüser C, et al.: Bax assembles into large ring-like structures remodeling the mitochondrial outer membrane in apoptosis. EMBO J. 2016; 35(4): 402–13. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nKasper R, Harke B, Forthmann C, et al.: Single-molecule STED microscopy with photostable organic fluorophores. Small. 2010; 6(13): 1379–84. PubMed Abstract | Publisher Full Text\n\nHanne J, Falk HJ, Görlitz F, et al.: STED nanoscopy with fluorescent quantum dots. Nat Commun. 2015; 6: 7127. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHeintzmann R, Cremer C: Axial tomographic confocal fluorescence microscopy. J Microsc. 2002; 206(Pt 1): 7–23. PubMed Abstract | Publisher Full Text\n\nBrown AC, Oddos S, Dobbie IM, et al.: Remodelling of cortical actin where lytic granules dock at natural killer cell immune synapses revealed by super-resolution microscopy. PLoS Biol. 2011; 9(9): e1001152. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGustafsson MG: Surpassing the lateral resolution limit by a factor of two using structured illumination microscopy. J Microsc. 2000; 198(Pt 2): 82–7. PubMed Abstract | Publisher Full Text\n\nRego EH, Shao L, Macklin JJ, et al.: Nonlinear structured-illumination microscopy with a photoswitchable protein reveals cellular structures at 50-nm resolution. Proc Natl Acad Sci U S A. 2012; 109(3): E135–43. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWatanabe K, Palonpon AF, Smith NI, et al.: Structured line illumination Raman microscopy. Nat Commun. 2015; 6: 10095. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nOldenburg J, van der Krogt G, Twiss F, et al.: VASP, zyxin and TES are tension-dependent members of Focal Adherens Junctions independent of the α-catenin-vinculin module. Sci Rep. 2015; 5: 17225. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFiolka R, Shao L, Rego EH, et al.: Time-lapse two-color 3D imaging of live cells with doubled resolution using structured illumination. Proc Natl Acad Sci U S A. 2012; 109(14): 5311–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu J, Min W, Conchello JA, et al.: Super-resolution laser scanning microscopy through spatiotemporal modulation. Nano Lett. 2009; 9(11): 3883–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWinter PW, Chandris P, Fischer RS, et al.: Incoherent structured illumination improves optical sectioning and contrast in multiphoton super-resolution microscopy. Opt Express. 2015; 23(4): 5327–34. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWinter PW, York AG, Nogare DD, et al.: Two-photon instant structured illumination microscopy improves the depth penetration of super-resolution imaging in thick scattering samples. Optica. 2014; 1(3): 181–91. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYork AG, Parekh SH, Dalle Nogare D, et al.: Resolution doubling in live, multicellular organisms via multifocal structured illumination microscopy. Nat Methods. 2012; 9(7): 749–54. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYork AG, Chandris P, Nogare DD, et al.: Instant super-resolution imaging in live cells and embryos via analog image processing. Nat Methods. 2013; 10(11): 1122–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nCurd A, Cleasby A, Makowska K, et al.: Construction of an instant structured illumination microscope. Methods. 2015; 88: 37–47. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFunatsu T, Harada Y, Higuchi H, et al.: Imaging and nano-manipulation of single biomolecules. Biophys Chem. 1997; 68(1–3): 63–72. PubMed Abstract | Publisher Full Text\n\nIshijima A, Kojima H, Funatsu T, et al.: Simultaneous observation of individual ATPase and mechanical events by a single myosin molecule during interaction with actin. Cell. 1998; 92(2): 161–71. PubMed Abstract | Publisher Full Text\n\nElting MW, Spudich JA: Future challenges in single-molecule fluorescence and laser trap approaches to studies of molecular motors. Dev Cell. 2012; 23(6): 1084–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakamoto T, Yildez A, Selvin PR, et al.: Step-size is determined by neck length in myosin V. Biochemistry. 2005; 44(49): 16203–10. PubMed Abstract | Publisher Full Text\n\nQu X, Wu D, Mets L, et al.: Nanometer-localized multiple single-molecule fluorescence microscopy. Proc Natl Acad Sci U S A. 2004; 101(31): 11298–303. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGordon MP, Ha T, Selvin PR: Single-molecule high-resolution imaging with photobleaching. Proc Natl Acad Sci U S A. 2004; 101(17): 6462–5. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChurchman LS, Okten Z, Rock RS, et al.: Single molecule high-resolution colocalization of Cy3 and Cy5 attached to macromolecules measures intramolecular distances through time. Proc Natl Acad Sci U S A. 2005; 102(5): 1419–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRust MJ, Bates M, Zhuang X: Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Methods. 2006; 3(10): 793–5. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBetzig E, Patterson GH, Sougrat R, et al.: Imaging intracellular fluorescent proteins at nanometer resolution. Science. 2006; 313(5793): 1642–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nAxelrod D: Total internal reflection fluorescence microscopy in cell biology. Methods Enzymol. 2003; 361: 1–33. PubMed Abstract | Publisher Full Text\n\nAxelrod D: Cell-substrate contacts illuminated by total internal reflection fluorescence. J Cell Biol. 1981; 89(1): 141–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShroff H, Galbraith CG, Galbraith JA, et al.: Live-cell photoactivated localization microscopy of nanoscale adhesion dynamics. Nat Methods. 2008; 5(5): 417–23. PubMed Abstract | Publisher Full Text\n\nRossy J, Pageon SV, Davis DM, et al.: Super-resolution microscopy of the immunological synapse. Curr Opin Immunol. 2013; 25(3): 307–12. PubMed Abstract | Publisher Full Text\n\nHess ST, Gould TJ, Gudheti MV, et al.: Dynamic clustered distribution of hemagglutinin resolved at 40 nm in living cell membranes discriminates between raft theories. Proc Natl Acad Sci U S A. 2007; 104(44): 17370–5. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRossi A, Moritz TJ, Ratelade J, et al.: Super-resolution imaging of aquaporin-4 orthogonal arrays of particles in cell membranes. J Cell Sci. 2012; 125(Pt 18): 4405–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGudheti MV, Curthoys NM, Gould TJ, et al.: Actin mediates the nanoscale membrane organization of the clustered membrane protein influenza hemagglutinin. Biophys J. 2013; 104(10): 2182–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaplante C, Huang F, Bewersdorf J, et al.: High-Speed Super-Resolution Imaging of Live Fission Yeast Cells. Methods Mol Biol. 2016; 1369: 45–57. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nShtengel G, Galbraith JA, Galbraith CG, et al.: Interferometric fluorescent super-resolution microscopy resolves 3D cellular ultrastructure. Proc Natl Acad Sci U S A. 2009; 106(9): 3125–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanchanawong P, Shtengel G, Pasapera AM, et al.: Nanoscale architecture of integrin-based cell adhesions. Nature. 2010; 468(7323): 580–4. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nZhong G, He J, Zhou R, et al.: Developmental mechanism of the periodic membrane skeleton in axons. eLife. 2014; 3: e04581. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSong AH, Wang D, Chen G, et al.: A selective filter for cytoplasmic transport at the axon initial segment. Cell. 2009; 136(6): 1148–60. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWatanabe K, Al-Bassam S, Miyazaki Y, et al.: Networks of polarized actin filaments in the axon initial segment provide a mechanism for sorting axonal and dendritic proteins. Cell Rep. 2012; 2(6): 1546–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetersen JD, Kaech S, Banker G: Selective microtubule-based transport of dendritic membrane proteins arises in concert with axon specification. J Neurosci. 2014; 34(12): 4135–47. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLeterrier C, Potier J, Caillol G, et al.: Nanoscale Architecture of the Axon Initial Segment Reveals an Organized and Robust Scaffold. Cell Rep. 2015; 13(12): 2781–93. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMennella V, Keszthelyi B, McDonald KL, et al.: Subdiffraction-resolution fluorescence microscopy reveals a domain of the centrosome critical for pericentriolar material organization. Nat Cell Biol. 2012; 14(11): 1159–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang B, Jones SA, Brandenburg B, et al.: Whole-cell 3D STORM reveals interactions between cellular structures with nanometer-scale resolution. Nat Methods. 2008; 5(12): 1047–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiedentopf H, Zsigmondy R: Uber Sichtbarmachung und Größenbestimmung ultramikoskopischer Teilchen, mit besonderer Anwendung auf Goldrubingläser. Ann Phys. 1902; 315: 1–39. Publisher Full Text\n\nHuisken J, Swoger J, Del Bene F, et al.: Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science. 2004; 305(5686): 1007–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nLin Y, Seka W, Eberly JH, et al.: Experimental investigation of Bessel beam characteristics. Appl Opt. 1992; 31(15): 2708–13. PubMed Abstract | Publisher Full Text\n\nFahrbach FO, Rohrbach A: A line scanned light-sheet microscope with phase shaped self-reconstructing beams. Opt Express. 2010; 18(23): 24229–44. PubMed Abstract | Publisher Full Text\n\nDurnin J, Miceli J Jr, Eberly JH: Diffraction-free beams. Phys Rev Lett. 1987; 58(15): 1499–501. PubMed Abstract | Publisher Full Text\n\nPlanchon TA, Gao L, Milkie DE, et al.: Rapid three-dimensional isotropic imaging of living cells using Bessel beam plane illumination. Nat Methods. 2011; 8(5): 417–23. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFahrbach FO, Gurchenkov V, Alessandri K, et al.: Light-sheet microscopy in thick media using scanned Bessel beams and two-photon fluorescence excitation. Opt Express. 2013; 21(11): 13824–39. PubMed Abstract | Publisher Full Text\n\nGao L, Shao L, Chen BC, et al.: 3D live fluorescence imaging of cellular dynamics using Bessel beam plane illumination microscopy. Nat Protoc. 2014; 9(5): 1083–101. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nZhao M, Zhang H, Li Y, et al.: Cellular imaging of deep organ using two-photon Bessel light-sheet nonlinear structured illumination microscopy. Biomed Opt Express. 2014; 5(5): 1296–308. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAdams MW, Loftus AF, Dunn SE, et al.: Light Sheet Fluorescence Microscopy (LSFM). Curr Protoc Cytom. 2015; 71: 12.37.1–15. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGao L, Shao L, Higgins CD, et al.: Noninvasive imaging beyond the diffraction limit of 3D dynamics in thickly fluorescent specimens. Cell. 2012; 151(6): 1370–85. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChen BC, Legant WR, Wang K, et al.: Lattice light-sheet microscopy: imaging molecules to embryos at high spatiotemporal resolution. Science. 2014; 346(6208): 1257998. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWu Y, Ghitani A, Christensen R, et al.: Inverted selective plane illumination microscopy (iSPIM) enables coupled cell identity lineaging and neurodevelopmental imaging in Caenorhabditis elegans. Proc Natl Acad Sci U S A. 2011; 108(43): 17708–13. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nKeller PJ, Schmidt AD, Wittbrodt J, et al.: Reconstruction of zebrafish early embryonic development by scanned light sheet microscopy. Science. 2008; 322(5904): 1065–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTomer R, Khairy K, Amat F, et al.: Quantitative high-speed imaging of entire developing embryos with simultaneous multiview light-sheet microscopy. Nat Methods. 2012; 9(7): 755–63. PubMed Abstract | Publisher Full Text\n\nKrzic U, Gunther S, Saunders TE, et al.: Multiview light-sheet microscope for rapid in toto imaging. Nat Methods. 2012; 9(7): 730–3. PubMed Abstract | Publisher Full Text\n\nWu Y, Wawrzusin P, Senseney J, et al.: Spatially isotropic four-dimensional imaging with dual-view plane illumination microscopy. Nat Biotechnol. 2013; 31(11): 1032–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar A, Wu Y, Christensen R, et al.: Dual-view plane illumination microscopy for rapid and spatially isotropic imaging. Nat Protoc. 2014; 9(11): 2555–73. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPitrone PG, Schindelin J, Stuyvenberg L, et al.: OpenSPIM: an open-access light-sheet microscopy platform. Nat Methods. 2013; 10(7): 598–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGualda EJ, Vale T, Almada P, et al.: OpenSpinMicroscopy: an open-source integrated microscopy platform. Nat Methods. 2013; 10(7): 599–600. PubMed Abstract | Publisher Full Text\n\nTomer R, Lovett-Barron M, Kauvar I, et al.: SPED Light Sheet Microscopy: Fast Mapping of Biological System Structure and Function. Cell. 2015; 163(7): 1796–806. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChung K, Wallace J, Kim SY, et al.: Structural and molecular interrogation of intact biological systems. Nature. 2013; 497(7449): 332–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation"
}
|
[
{
"id": "14733",
"date": "30 Jun 2016",
"name": "Hari Shroff",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14734",
"date": "30 Jun 2016",
"name": "Gawain McColl",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1553
|
https://f1000research.com/articles/5-1552/v1
|
30 Jun 16
|
{
"type": "Review",
"title": "Urinary tract infection in the setting of vesicoureteral reflux",
"authors": [
"Michael L. Garcia-Roig",
"Andrew J. Kirsch",
"Michael L. Garcia-Roig"
],
"abstract": "Vesicoureteral reflux (VUR) is the most common underlying etiology responsible for febrile urinary tract infections (UTIs) or pyelonephritis in children. Along with the morbidity of pyelonephritis, long-term sequelae of recurrent renal infections include renal scarring, proteinuria, and hypertension. Treatment is directed toward the prevention of recurrent infection through use of continuous antibiotic prophylaxis during a period of observation for spontaneous resolution or by surgical correction. In children, bowel and bladder dysfunction (BBD) plays a significant role in the occurrence of UTI and the rate of VUR resolution. Effective treatment of BBD leads to higher rates of spontaneous resolution and decreased risk of UTI.",
"keywords": [
"Vesicoureteral reflux",
"urinary tract infections",
"pyelonephritis",
"pediatric urology"
],
"content": "Introduction\n\nThe prevalence of febrile urinary tract infection (UTI) in infants and children ranges from 3 to 7% and varies by age, race, sex, and circumcision status1–3. Vesicoureteral reflux (VUR) is found in 30–45% of children presenting with a febrile UTI, with an even higher risk in neonates4,5. UTI in the setting of VUR is often associated with pyelonephritis, as reflux results in direct communication of infected urine between the bladder and kidney, thus permitting cystitis to rapidly progress to acute pyelonephritis.\n\nThe diagnosis and management of VUR have been met with controversy between recent UTI guidelines published by the American Academy of Pediatrics (AAP) and the American Urological Association (AUA) reflux guidelines. Several studies have shown no significant benefit to continuous antibiotic prophylaxis (CAP) in the lowest grades of VUR, favoring a less aggressive screening and treatment algorithm. In such cases, a period of observation off CAP has been recommended. In contrast, recent large randomized controlled trials of mild to severe VUR have shown CAP and/or anti-reflux surgery to minimize recurrent pyelonephritis and, in some reports, renal scarring6,7. VUR of any grade should be viewed as one of many risk factors increasing the chance of recurrent pyelonephritis. We outline the recent advances in the medical and surgical management of UTI in the setting of VUR.\n\n\nThe impact of UTI on VUR\n\nAcquired renal scarring associated with VUR is the result of the acute inflammatory reaction that develops secondary to bacterial infection of the renal parenchyma. The inflammation is mediated by cytokine release, resulting in focal parenchymal ischemia and, ultimately, scarring8. Refluxing sterile urine is not a detriment to renal function, but high-grade VUR is often associated with congenital renal dysplasia and can affect renal architecture. The extent of renal damage after a febrile infection depends on bacterial and host factors that mediate the response to infection. Late sequelae of renal scarring, such as hypertension, proteinuria, or even chronic renal failure, can be seen in the second or third decades of life9.\n\n\nA shift toward observation with CAP for selected patients\n\nVUR diagnosed in early childhood has been found to resolve spontaneously and safely in some patients after a period of observation on CAP. Prognostic calculators have been developed to predict VUR resolution in children10–12. Unfortunately, the international VUR grading system alone is subject to misinterpretation by radiologists and is over-simplistic regarding the many nuances of VUR grading. Our group recently validated a six-point scoring system based on voiding cystourethrogram (VCUG) findings that accurately predicts the likelihood of VUR resolution in children younger than 2 years old based on gender, VUR grade, VUR timing, and ureteral abnormalities10. Our data, as well as others’, emphasize that low-pressure VUR (i.e. VUR occurring at low bladder volume) has a significantly lower rate of resolution when compared to the same grade of VUR occurring later in the bladder cycle (i.e. late filling or voiding VUR)10,12–14. These recent developments provide reasonable expectations for spontaneous VUR resolution to the provider and parent based on limited imaging and patient characteristics.\n\nThe 2010 AUA’s guideline on VUR recommended goals of VUR treatment as preventing recurrent febrile UTI and renal scarring and minimizing the morbidity of treatment15. A period of observation for VUR resolution is indicated given that these goals can be met. CAP has become a mainstay during this waiting period, as it reduces the rate of febrile UTI in the setting of VUR. This evidence came, in part, from the results of the Swedish reflux and the Randomized Intervention for Children with Vesicoureteral Reflux (RIVUR) trials7,15. The Swedish reflux trial randomized 203 children with grade III-IV VUR to placebo, CAP, or endoscopic injection and reported a recurrent UTI rate in girls of 19% on prophylaxis, 23% with endoscopic injection, and 57% on surveillance (p = 0.0002), while no difference was observed in boys6. The RIVUR trial randomized 607 children with VUR diagnosed after febrile UTI to placebo vs. CAP and found a 50% reduction in the risk of UTI recurrence in those on prophylaxis7. A subsequent meta-analysis confirmed the benefit of CAP in reducing the risk of febrile or symptomatic UTI in children with VUR (pooled odds ratio [OR] 0.63, 95% confidence interval [CI] 0.42–0.96)16. A link between CAP and reduction of renal scarring is less well defined than between CAP and UTI reduction, as the RIVUR trial (inclusive of grades I-IV) and the recent meta-analysis both reported CAP use was not associated with a decrease in new renal scarring, while the Swedish reflux trial (inclusive of grades III-IV) showed a reduction in renal scarring for the CAP group6,16,17. It should be emphasized that the genesis of renal scarring may take years to detect on renal imaging studies and that the primary outcome of the RIVUR trial was not to show a reduction in renal scarring.\n\n\nRisk factors for breakthrough UTI\n\nAlthough CAP has been shown to reduce the rate of recurrent UTI, risks of recurrent pyelonephritis are not inconsequential. Factors influencing the rate of breakthrough UTI in children on CAP include a multitude of factors, as illustrated in Figure 1. Bowel and bladder dysfunction (BBD) in the setting of VUR results in a 56% risk of recurrent UTI vs. 25.4% in children with VUR only18. Antibiotic prophylaxis in patients with BBD and VUR is particularly effective at reducing the risk of recurrent UTI (hazard ratio [HR] 0.21, 95% CI, 0.08–0.58)7. During a period of observation, patients with dysfunctional voiding should undergo targeted BBD treatment, as this increases the likelihood of spontaneous VUR resolution (70%) compared to those with idiopathic detrusor overactivity (38%) or detrusor underutilization (40%) in patients with mild to moderate VUR19.\n\nAs shown, vesicoureteral reflux (VUR) is one of several important risk factors illustrating the multifactorial nature of urinary tract infection (UTI)/VUR management. Individual factors may or may not be present in an individual patient and play varying roles in UTI recurrence and VUR resolution and management.\n\nAs stated above, the timing of VUR during a VCUG is an important prognostic marker for its resolution and risk of breakthrough febrile UTI. Alexander et al. identified that VUR onset at ≤35% bladder capacity on VCUG was an independent predictor of breakthrough UTI (HR 1.58, 95% CI 1.05–2.38, p = 0.03)13. The likelihood of VUR resolution is also correlated with bladder volume at onset of VUR, where VUR at >50% predicted bladder capacity is more likely to resolve spontaneously (p<0.001)10,11,14. In our recent studies, early filling VUR has been shown to be the most important predictor of non-resolution in children diagnosed before 2 years of age when calculating risk based on the VUR index10,11. The VUR index and its associated resolution rate are outlined in Figure 2. These studies taken together suggest that early filling, low-volume VUR, despite its grade, predicts both non-resolution and increased risk of acute pyelonephritis and should be the focus of further research regarding VUR management.\n\nThe rate of resolution or improvement is outlined in the graph of improvement rate based on VUR index score. The graph represents an average of the initial VUR index cohort and subsequent multi-institutional VUR index validation cohort.\n\nParent adherence to CAP is a concern in patients with VUR, as CAP is effective only if it is given consistently. Medication adherence rates in the setting of chronic disease are relatively low, reported at 50–70%20–22. The RIVUR study documented administration of CAP at least 50% of the time in 85.2% of patients and 75% of the time in 76.7% based on parent questionnaire7. However, Smyth et al. and Eanaretto et al. painted a more dismal picture regarding CAP adherence by testing urine for the presence of antibiotics with a urine positive rate of 17–67%23,24. This rate of medication adherence reflects the challenge physicians face in treating not only the disease but also the patient. The ultimate approach to VUR treatment, whether medical or surgical, involves an informed discussion with parents and that weighs all available options.\n\n\nAntibiotic choice for breakthrough UTI\n\nAntibiotics are the mainstay of treatment for bacterial UTIs. Bacterial cystitis is more likely to predispose to pyelonephritis in the setting of VUR. Treatment is geared toward relief of symptoms through killing the offending bacterial pathogen with antibiotics. Management centers on long-established principles of antibiotic use, including confirming the presence of infection with urine analysis followed by culture and sensitivity testing and administering empiric antibiotics for common uropathogens based on local antibiograms, or previous positive urine cultures if present, and a culture-specific antibiotic once antibiotic sensitivity is available25.\n\nIn the context of VUR, antibiotics are used effectively as prophylaxis against bacterial infection due to the risk of renal scarring and permanent secondary renal damage26. The downstream effect of this is an increased presence of bacteria resistant to the antibiotic given in the urinary and gastrointestinal tract7,27. The temporal relationship of antibiotic exposure matters in the presence of resistant bacteria. After studying 500 children presenting with an initial UTI, Paschke and colleagues demonstrated a fourfold increased risk (OR 3.6, 95% CI 1.6–8.2) of ampicillin or amoxicillin/clavulanic acid resistance among pathogens with amoxicillin exposure in the 30 days prior to UTI; however, antibiotic exposure >60 days prior to UTI did not portend presence of ampicillin-resistant bacteria28. Other authors have highlighted that antibiotic resistance occurs at a much faster rate than the decay of resistance over time29. These factors argue in favor of careful consideration of targeted antibiotic choice for empiric treatment and culture-specific prophylaxis in the setting of patients with potential prolonged antibiotic exposure.\n\nThe role of prophylaxis after UTI may relate to the grade of VUR present. Several studies have concluded that prophylaxis is less beneficial in preventing UTI in children with grade I-III VUR because the baseline risk of infection is very low30. However, in children with high-grade reflux, grade IV-V, the risk of UTI recurrence is increased fourfold and prophylaxis does offer significant benefit30,31. It should be emphasized that the majority of studies showing no benefit of CAP in patients with low-grade VUR included mostly grades I-II. Moderate grade III VUR has been shown to be associated with a higher risk of UTI and renal scarring compared to lower VUR grades. For example, the RIVUR trial study population consisted of 8% with grade IV VUR and no patients with grade V. Although children with grade I-II VUR were at a lower baseline risk of VUR, a reduction in recurrent UTI was noted in all groups regardless of VUR grade7. Confounding the issue regarding VUR grade is the nearly 33% chance that radiologists will misgrade mild to moderate VUR and adjudication of VCUGs most often results in a higher reported grade7. Furthermore, studies comparing VUR to no VUR on a single VCUG that show no difference in UTI risk may be including many patients who actually have VUR but classified otherwise, since a non-cyclic VCUG may be falsely negative in up to 20% of cases32. Furthermore, the diagnosis and treatment of occult VUR (i.e. recurrent pyelonephritis despite a “normal” VCUG) has been shown to significantly reduce the incidence of acute pyelonephritis33.\n\n\nThe role of anti-reflux surgery in reducing breakthrough UTI\n\nElimination of reflux by surgical means is an effective approach to treatment. Indications for surgical correction include breakthrough UTI while on CAP, poor adherence to CAP resulting in infection, persistence of VUR after a period of observation, low likelihood of spontaneous resolution in a high-risk patient, or parent’s preference given the benefits and risks of each treatment modality. Surgical options include open or laparoscopic ureteral reimplantation or endoscopic injection of a bulking agent (dextranomer/hyaluronic acid copolymer, Deflux®, is currently the only Food and Drug Administration-approved agent in the USA). Success after these procedures is defined clinically and radiographically, with clinical success being the absence of recurrent UTI after cessation of CAP and radiographically by the absence of VUR on postoperative VCUG. Radiographic success for open ureteral reimplantation is 95-98% for primary low- to moderate-grade VUR and 94% for higher-grade VUR34,35. Surgical treatment success of grade V VUR is approximately 80%. Clinical success after open surgery has been reported to be between 80 and 95% and corresponds to the rate of preoperative UTIs and presence of renal involvement on a dimercaptosuccinic acid (DMSA) renal scan or sonography36–38.\n\nRobot-assisted ureteroneocystostomy is a relatively new procedure compared to the open approaches and fewer reports are available, with clinical success recently reported at 93% by one center37 and radiographic success at 77–92%37,38. Our experience with robot-assisted extravesical reimplantation has been favorable, with greater than 90% clinical and radiographic success in a complex patient cohort including reoperative surgery for VUR and obstruction39. The best approach for the patient, whether open or robot assisted, depends on patient and parent preference after a discussion of the pros and cons of each approach, such as surgical scar location and postoperative convalescence.\n\nEndoscopic injection of dextranomer/hyaluronic acid into the subureteric space is an additional method of minimally invasive treatment for VUR. Reported clinical and radiographic success ranges from 50–93%, prompting some surgeons to avoid this method of treatment in favor of open or robot-assisted approaches8,40. The success rate is impacted by a learning curve with injection, as demonstrated by Lee et al., who reported an improvement from 65.9% in their first 337 ureters to 80.2% in a follow up group41. Our group has demonstrated modifications in technique that resulted in a consistent 90% radiographic and 93% clinical success in children with primary grades I-IV VUR. Complex cases, such as duplex ureters or injection following failed open surgery, tend to have an approximately 10% lower success. With the double hydrodistention implantation technique (HIT), we specifically emphasize the importance of a minimum 1 ml per ureter volume of substance injected in two tandem injection points to coapt the ureteral tunnel and orifice and an objective end-point of injection being the absence of ureteral hydrodistention42,43.\n\nAfter surgical correction of VUR, patients are maintained on prophylactic antibiotics until the absence of hydronephrosis on ultrasound is confirmed approximately 4–6 weeks after surgery. This protocol is followed independent of approach. Repeat VCUG is not performed at our institution after surgical correction with open or endoscopic techniques owing to the high documented success rate with open and endoscopic procedures44–46. However, postoperative VCUG is currently performed after robot-assisted ureteroneocystostomy due to the relatively new nature of this procedure. Postoperatively, patients are followed clinically for signs of infection after renal ultrasound confirms stable renal appearance and prophylactic antibiotics are stopped.\n\n\nBreakthrough UTI after anti-reflux surgery\n\nPersistent reflux or recurrent UTI are possible after open or endoscopic surgical VUR correction, representing radiographic or clinical failure, respectively. The risk for these varies based on procedure type, as outlined above. Identification of persistent low-grade VUR after ureteroneocystostomy does not necessarily correlate with a risk of postoperative breakthrough UTI, independent of procedure type35,47. Once a patient is treated for the UTI recurrence, prophylaxis is continued until definitive treatment can be performed to avoid a recurrent breakthrough UTI. BBD should be addressed if present, as this is an independent and treatable risk factor for UTI both before and after anti-reflux surgery19.\n\nThe major indication for revision surgery after corrective VUR surgery falls on the occurrence of a breakthrough febrile UTI. The surgical approach to revision surgery depends on surgeon preference. Recurrent UTI after open or robot-assisted ureteroneocystostomy can be successfully managed endoscopically or by repeat open or robot-assisted ureteroneocystostomy. Perez-Brayfield et al. reported the success of endoscopic injection at 88% in patients with persistent VUR after open ureteral reimplantation48. Two small studies by Kitchens and Jung documented the utility of endoscopic injection following open ureteroneocystostomy, including previous cross-trigonal and extravesical approaches, with resolution occurring after one injection in 70–83%49,50.\n\nThe largest series reporting success after endoscopic re-injection is from Puri et al., with 1551 children (2341 ureters) undergoing primary endoscopic injection for grades II-V VUR (92.2% grade III-IV) and resolution occurring after a single injection in 87.1%, second injection in 11.3%, and third in 1.6%51. The indication for repeat injection was persistent VUR, which was more common in younger children and those with a higher grade of VUR (p<0.001). Persistent VUR after repeat injection is managed by ureteroneocystostomy, which can be performed via an open approach or with robot assistance.\n\nRobot-assisted ureteroneocystostomy in children with persistent VUR after previous anti-reflux surgery is quickly becoming a mainstay of treatment. Arlen et al. reported the success after robot-assisted ureteral reimplant in 11 previously treated children (10 endoscopic injection and one open reimplant) with complete resolution in all reimplanted ureters; however, one developed new onset contralateral VUR39.\n\n\nConclusion and future considerations\n\nA shift toward observation with CAP for young patients with VUR provides an option for conservative management. Breakthrough UTI represents potential for renal damage and scarring for these patients and those who underwent surgical correction of VUR. Risk factors for breakthrough UTI and spontaneous VUR resolution have been identified, as outlined above. Future considerations for improved management of this disease include less invasive VUR diagnosis methods and non-antibiotic alternatives to UTI prophylaxis.\n\nCurrent recommendations for VUR diagnosis after febrile UTI have excluded the VCUG due to its invasive nature and potential for iatrogenic UTI52. The 2011 AAP UTI guidelines recommend evaluation with a renal ultrasound after the first febrile UTI and VCUG after the second. These guidelines have resulted in a substantial decrease in VCUGs and have potentially limited the use of sonograms as well53. A particular concern of limiting VUR diagnosis is that many children with BBD or other urological issues will not be referred to specialists who may offer effective treatments for VUR and voiding dysfunction. While a non-invasive method for accurate VUR diagnosis would prevent patients and providers from waiting until a second episode of pyelonephritis before diagnosis54, such technology is currently not readily available. Despite its many disadvantages, the VCUG remains the only test available to reliably diagnose and grade VUR.\n\nA period of observation for spontaneous VUR resolution relies on CAP to prevent UTI recurrence and the morbidity of acute pyelonephritis and renal damage. Several arguments exist against CAP, specifically its encouragement of the development of drug-resistant bacteria and the unclear impact of antibiotics on host flora. Some reports suggest that non-antibiotic management of bacterial cystitis is effective; however, a consistent method does not exist for UTI prevention or pyelonephritis55,56. Current evidence supports traditional methods of UTI and VUR diagnosis and should continue to form the cornerstone of management until less invasive management options become available.\n\n\nAbbreviations\n\nAAP, American Academy of Pediatrics; AUA, American Urological Association; BBD, bowel and bladder dysfunction; CAP, continuous antibiotic prophylaxis; CI, confidence interval; DMSA, dimercaptosuccinic acid; HIT, hydrodistention implantation technique; HR, hazard ratio; OR, odds ratio; RIVUR, Randomized Intervention for Children with Vesicoureteral Reflux; UTI, urinary tract infection; VCUG, voiding cystourethrogram; VUR, vesicoureteral reflux",
"appendix": "Competing interests\n\n\n\nThe author(s) declare that they have no disclosures.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHoberman A, Chao HP, Keller DM, et al.: Prevalence of urinary tract infection in febrile infants. J Pediatr. 1993; 123(1): 17–23. PubMed Abstract | Publisher Full Text\n\nShaikh N, Morone NE, Bost JE, et al.: Prevalence of urinary tract infection in childhood: a meta-analysis. Pediatr Infect Dis J. 2008; 27(4): 302–8. PubMed Abstract | Publisher Full Text\n\nShaw KN, Gorelick M, McGowan KL, et al.: Prevalence of urinary tract infection in febrile young children in the emergency department. Pediatrics. 1998; 102(2): e16. PubMed Abstract\n\nHiraoka M, Hori C, Tsukahara H, et al.: Vesicoureteral reflux in male and female neonates as detected by voiding ultrasonography. Kidney Int. 1999; 55(4): 1486–90. PubMed Abstract | Publisher Full Text\n\nHoberman A, Charron M, Hickey RW, et al.: Imaging studies after a first febrile urinary tract infection in young children. N Engl J Med. 2003; 348(3): 195–202. PubMed Abstract | Publisher Full Text\n\nBrandström P, Jodal U, Sillén U, et al.: The Swedish reflux trial: review of a randomized, controlled trial in children with dilating vesicoureteral reflux. J Pediatr Urol. 2011; 7(6): 594–600. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRIVUR Trial Investigators; Hoberman A, Greenfield SP, et al.: Antimicrobial prophylaxis for children with vesicoureteral reflux. N Engl J Med. 2014; 370(25): 2367–76. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPeters C, Rushton HG: Vesicoureteral reflux associated renal damage: congenital reflux nephropathy and acquired renal scarring. J Urol. 2010; 184(1): 265–73. PubMed Abstract | Publisher Full Text\n\nJacobson SH, Eklöf O, Eriksson CG, et al.: Development of hypertension and uraemia after pyelonephritis in childhood: 27 year follow up. BMJ. 1989; 299(6701): 703–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArlen AM, Garcia-Roig M, Weiss AD, et al.: Vesicoureteral Reflux Index: 2-Institution Analysis and Validation. J Urol. 2016; 195(4 Pt 2): 1294–9. PubMed Abstract | Publisher Full Text\n\nKirsch AJ, Arlen AM, Leong T, et al.: Vesicoureteral reflux index (VURx): a novel tool to predict primary reflux improvement and resolution in children less than 2 years of age. J Pediatr Urol. 2014; 10(6): 1249–54. PubMed Abstract | Publisher Full Text\n\nShiraishi K, Matsuyama H, Nepple KG, et al.: Validation of a prognostic calculator for prediction of early vesicoureteral reflux resolution in children. J Urol. 2009; 182(2): 687–90; discussion 690–1. PubMed Abstract | Publisher Full Text\n\nAlexander SE, Arlen AM, Storm DW, et al.: Bladder volume at onset of vesicoureteral reflux is an independent risk factor for breakthrough febrile urinary tract infection. J Urol. 2015; 193(4): 1342–6. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKnudson MJ, Austin JC, McMillan ZM, et al.: Predictive factors of early spontaneous resolution in children with primary vesicoureteral reflux. J Urol. 2007; 178(4 Pt 2): 1684–8. PubMed Abstract | Publisher Full Text\n\nPeters CA, Skoog SJ, Arant BS Jr, et al.: Summary of the AUA Guideline on Management of Primary Vesicoureteral Reflux in Children. J Urol. 2010; 184(3): 1134–44. PubMed Abstract | Publisher Full Text\n\nWang HH, Gbadegesin RA, Foreman JW, et al.: Efficacy of antibiotic prophylaxis in children with vesicoureteral reflux: systematic review and meta-analysis. J Urol. 2015; 193(3): 963–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMattoo TK, Chesney RW, Greenfield SP, et al.: Renal Scarring in the Randomized Intervention for Children with Vesicoureteral Reflux (RIVUR) Trial. Clin J Am Soc Nephrol. 2016; 11(1): 54–61. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nKeren R, Shaikh N, Pohl H, et al.: Risk Factors for Recurrent Urinary Tract Infection and Renal Scarring. Pediatrics. 2015; 136(1): e13–21. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFast AM, Nees SN, van Batavia JP, et al.: Outcomes of targeted treatment for vesicoureteral reflux in children with nonneurogenic lower urinary tract dysfunction. J Urol. 2013; 190(3): 1028–32. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nFesta RS, Tamaroff MH, Chasalow F, et al.: Therapeutic adherence to oral medication regimens by adolescents with cancer. I. Laboratory assessment. J Pediatr. 1992; 120(5) 807–11. PubMed Abstract | Publisher Full Text\n\nJay S, Litt IF, Durant RH: Compliance with therapeutic regimens. J Adolesc Health Care. 1984; 5(2): 124–36. PubMed Abstract | Publisher Full Text\n\nRapoff MA: Compliance with treatment regimens for pediatric rheumatic diseases. Arthritis Care Res. 1989; 2(3): S40–7. PubMed Abstract | Publisher Full Text\n\nPanaretto K, Craig J, Knight J, et al.: Risk factors for recurrent urinary tract infection in preschool children. J Paediatr Child Health. 1999; 35(5): 454–9. PubMed Abstract | Publisher Full Text\n\nSmyth AR, Judd BA: Compliance with antibiotic prophylaxis in urinary tract infection. Arch Dis Child. 1993; 68(2): 235–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdlin RS, Copp HL: Antibiotic resistance in pediatric urology. Ther Adv Urol. 2014; 6(2): 54–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoulthard MG, Lambert HJ, Keir MJ: Occurrence of renal scars in children after their first referral for urinary tract infection. BMJ. 1997; 315(7113): 918–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCostelloe C, Metcalfe C, Lovering A, et al.: Effect of antibiotic prescribing in primary care on antimicrobial resistance in individual patients: systematic review and meta-analysis. BMJ. 2010; 340: c2096. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPaschke AA, Zaoutis T, Conway PH, et al.: Previous antimicrobial exposure is associated with drug-resistant urinary tract infections in children. Pediatrics. 2010; 125(4): 664–72. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nAustin DJ, Kristinsson KG, Anderson RM: The relationship between the volume of antimicrobial consumption in human communities and the frequency of resistance. Proc Natl Acad Sci U S A. 1999; 96(3): 1152–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConway PH, Cnaan A, Zaoutis T, et al.: Recurrent urinary tract infections in children: risk factors and association with prophylactic antimicrobials. JAMA. 2007; 298(2): 179–86. PubMed Abstract | Publisher Full Text\n\nBrandström P, Esbjörner E, Herthelius M, et al.: The Swedish reflux trial in children: III. Urinary tract infection pattern. J Urol. 2010; 184(1): 286–91. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPapadopoulou F, Efremidis SC, Oiconomou A, et al.: Cyclic voiding cystourethrography: is vesicoureteral reflux missed with standard voiding cystourethrography? Eur Radiol. 2002; 12(3): 666–70. PubMed Abstract | Publisher Full Text\n\nArlen AM, Broderick KM, Leong T, et al.: Effective endoscopic diagnosis and treatment of pediatric occult vesicoureteral reflux with intermediate to long-term follow-up. J Pediatr Urol. 2014; 10(6): 1095–9. PubMed Abstract | Publisher Full Text\n\nBarrieras D, Lapointe S, Reddy PP, et al.: Are postoperative studies justified after extravescial ureteral reimplantation? J Urol. 2000; 164(3 pt 2): 1064–6. PubMed Abstract\n\nGrossklaus DJ, Pope JC, Adams MC, et al.: Is postoperative cystography necessary after ureteral reimplantation? Urology. 2001; 58(6): 1041–5. PubMed Abstract | Publisher Full Text\n\nElmore JM, Kirsch AJ, Heiss EA, et al.: Incidence of urinary tract infections in children after successful ureteral reimplantation versus endoscopic dextranomer/hyaluronic acid implantation. J Urol. 2008; 179(6): 2364–7: discussion 2367–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nAkhavan A, Avery D, Lendvay TS: Robot-assisted extravesical ureteral reimplantation: outcomes and conclusions from 78 ureters. J Pediatr Urol. 2014; 10(5): 864–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGrimsby GM, Dwyer ME, Jacobs MA, et al.: Multi-institutional review of outcomes of robot-assisted laparoscopic extravesical ureteral reimplantation. J Urol. 2015; 193(5 Suppl): 1791–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nArlen AM, Broderick KM, Travers C, et al.: Outcomes of complex robot-assisted extravesical ureteral reimplantation in the pediatric population. J Pediatr Urol. 2015; pii: S1477-5131(15)00445-3. PubMed Abstract | Publisher Full Text\n\nRouth JC, Inman BA, Reinberg Y: Dextranomer/hyaluronic acid for pediatric vesicoureteral reflux: systematic review. Pediatrics. 2010; 125(5): 1010–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nLee EK, Gatti JM, Demarco RT, et al.: Long-term followup of dextranomer/hyaluronic acid injection for vesicoureteral reflux: late failure warrants continued followup. J Urol. 2009; 181(4): 1869–74: discussion 1874–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKalisvaart JF, Scherz HC, Cuda S, et al.: Intermediate to long-term follow-up indicates low risk of recurrence after Double HIT endoscopic treatment for primary vesico-ureteral reflux. J Pediatr Urol. 2012; 8(4): 359–65. PubMed Abstract | Publisher Full Text\n\nKaye JD, Srinivasan AK, Delaney C, et al.: Clinical and radiographic results of endoscopic injection for vesicoureteral reflux: defining measures of success. J Pediatr Urol. 2012; 8(3): 297–303. PubMed Abstract | Publisher Full Text\n\nArlen AM, Scherz HC, Filimon E, et al.: Is routine voiding cystourethrogram necessary following double hit for primary vesicoureteral reflux? J Pediatr Urol. 2015; 11(1): 40.e1–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBisignani G, Decter RM: Voiding cystourethrography after uncomplicated ureteral reimplantation in children: is it necessary? J Urol. 1997; 158(3 Pt 2): 1229–31. PubMed Abstract | Publisher Full Text\n\nBomalaski MD, Ritchey ML, Bloom DA: What imaging studies are necessary to determine outcome after ureteroneocystostomy? J Urol. 1997; 158(3 Pt 2): 1226–8. PubMed Abstract | Publisher Full Text\n\nDwyer ME, Husmann DA, Rathbun SR, et al.: Febrile urinary tract infections after ureteroneocystostomy and subureteral injection of dextranomer/hyaluronic acid for vesicoureteral reflux--do choice of procedure and success matter? J Urol. 2013; 189(1): 275–82. PubMed Abstract | Publisher Full Text\n\nPerez-Brayfield M, Kirsch AJ, Hensle TW, et al.: Endoscopic treatment with dextranomer/hyaluronic acid for complex cases of vesicoureteral reflux. J Urol. 2004; 172(4 Pt 2): 1614–6. PubMed Abstract | Publisher Full Text\n\nJung C, Demarco RT, Lowrance WT, et al.: Subureteral injection of dextranomer/hyaluronic acid copolymer for persistent vesicoureteral reflux following ureteroneocystostomy. J Urol. 2007; 177(1): 312–5. PubMed Abstract | Publisher Full Text\n\nKitchens D, Minevich E, DeFoor W, et al.: Endoscopic injection of dextranomer/hyaluronic acid copolymer to correct vesicoureteral reflux following failed ureteroneocystostomy. J Urol. 2006; 176(4 Pt 2): 1861–3. PubMed Abstract | Publisher Full Text\n\nPuri P, Kutasy B, Colhoun E, et al.: Single center experience with endoscopic subureteral dextranomer/hyaluronic acid injection as first line treatment in 1,551 children with intermediate and high grade vesicoureteral reflux. J Urol. 2012; 188(4 Suppl): 1485–9. PubMed Abstract | Publisher Full Text\n\nSubcommittee on Urinary Tract Infection, Steering Committee on Quality Improvement and Management, Roberts KB: Urinary tract infection: clinical practice guideline for the diagnosis and management of the initial UTI in febrile infants and children 2 to 24 months. Pediatrics. 2011; 128(3): 595–610. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nArlen AM, Merriman LS, Kirsch JM, et al.: Early effect of American Academy of Pediatrics Urinary Tract Infection Guidelines on radiographic imaging and diagnosis of vesicoureteral reflux in the emergency room setting. J Urol. 2015; 193(5 Suppl): 1760–5. PubMed Abstract | Publisher Full Text\n\nSnow BW, Arunachalam K, De Luca V, et al.: Non-invasive vesicoureteral reflux detection: heating risk studies for a new device. J Pediatr Urol. 2011; 7(6): 624–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGágyor I, Bleidorn J, Kochen MM, et al.: Ibuprofen versus fosfomycin for uncomplicated urinary tract infection in women: randomised controlled trial. BMJ. 2015; 351: h6544. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTewary K, Narchi H: Recurrent urinary tract infections in children: Preventive interventions other than prophylactic antibiotics. World J Methodol. 2015; 5(2): 13–9. PubMed Abstract | Free Full Text"
}
|
[
{
"id": "14731",
"date": "30 Jun 2016",
"name": "Dominic C. Frimberger",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14732",
"date": "30 Jun 2016",
"name": "Martin Koyle",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1552
|
https://f1000research.com/articles/5-1550/v1
|
30 Jun 16
|
{
"type": "Research Article",
"title": "Blood loss predictive factors and transfusion practice during percutaneous nephrolithotomy of kidney stones: a prospective study",
"authors": [
"Firtantyo Adi Syahputra",
"Ponco Birowo",
"Nur Rasyid",
"Faisal Abdi Matondang",
"Endrika Noviandrini",
"Maruto Harjanggi Huseini",
"Nur Rasyid",
"Faisal Abdi Matondang",
"Endrika Noviandrini",
"Maruto Harjanggi Huseini"
],
"abstract": "Objectives Bleeding is the most common complication of percutaneous nephrolithotomy (PCNL). Injudicious transfusion is frequently performed in current practice, even though it is not always needed. This study aimed to identify the predictive factors of blood loss in the PCNL procedure and evaluate the perioperative transfusion practice.\n\nMethods A prospective study of PCNL was randomly performed by two consultants of endo-urology at our institution. The inclusion criteria were adults with kidney pelvic stones >20 mm or stone in inferior calyx >10 mm or staghorn stone. Those with coagulopathy, under anti-coagulant treatment or open conversion were excluded. A full blood count was taken at baseline and during 12, 24, 36, 72-hours post-operatively. Factors such as stone burden, sex, body surface area, shifting of hematocrit level and amount of blood transfused were analyzed statistically using line regression to identify the predictive factors of total blood loss (TBL).\n\nResults Eighty-five patients were enrolled in this study. Mean TBL was 560.92 ± 428.43 mL for both endo-urology surgeons. Stone burden was the most influential factor for TBL (p=0.037). Our results revealed that TBL (mL) = -153.379 + 0.229 × stone burden (mm2) + 0.203 x baseline serum hematocrit (%); thus considerably predicted the need for blood transfusion. A total of 87.1% patients did not receive perioperative transfusion, 3.5% received intra-operative transfusion, 7.1% received post-operative transfusion, 23% had both intra and post-operative transfusion, resulting in a cross-matched transfusion ratio of 7.72. Mean perioperative blood transfused was 356.00 ± 145.88 mL.",
"keywords": [
"Bleeding",
"nephrolithiasis",
"PCNL",
"transfusion"
],
"content": "Introduction\n\nKidney stones prove to be a common affliction in many countries worldwide because of high incidence and prevalence. In America, kidney stone incidence was found in 116 out of 100,000 individuals1. A higher incidence was discovered in the German population aged 14 years and older, amounting to 720 out of 100,000 individuals2. An excessively high number of cases was found in Asian countries, namely, 114.3 per 100,000 in Japan, while Iranian urolithiasis incidence was assessed at 145.1 in 20053. A global increase in kidney stone cases was determined in individuals of all ages, sex, and races4. In our institution we have, to date, treated an increasing number of kidney stone patients, from 182 in 1997 to 847 in 20025.\n\nPercutaneous nephrolithotomy (PCNL) is a urological minimally-invasive procedure to extract kidney stones by means of percutaneous access6. Nowadays, PCNL is widely accepted to treat those complex kidney stone cases that are hard in consistency, of a large size, infected, obstructed with anatomical abnormalities, generally cases that could not be treated by other modalities7. In European guidelines, PCNL is favoured as the treatment of choice for calyx and pelvic stones >20 mm28. For stones >20 mm, PCNL demonstrated a stone-free rate up to 78–95%9. There has been a decrease in open surgery number in our institution during 2001–2009, whereby the PCNL procedure has become more frequent10.\n\nOne of the most bothersome complications of PCNL is hemorrhage. Direct access to the pelvicalyceal system and intrarenal manipulation during PCNL procedures cause injury to the renal vasculature, particularly to the segmental and interlobar arteries. The renal-collecting system is rich in vascularization, covering 20% of the total cardiac output, and often results in hemorrhage during PCNL11. The high percentage of blood loss and the necessity of transfusion often results in the erroneous management of hemorrhage during the PCNL procedure.\n\nIt has been reported that 1–11% of patients who underwent PCNL required blood transfusion; a higher transfusion rate, 2–53% was determined in the staghorn cases11. Previous studies describe many hemorrhage risk factors during PCNL, such as, age12, pre-operative urinary tract infections12, large stones (exceeding 1250 mm2)13, staghorn calculi13,14, multiple access11,13,14, diabetes mellitus11,13,15, prolonged surgery time11–13 and stone composition13. Other risk factors have been postulated: i.e. stone location, pre-operative hemoglobin, hydronephrosis grade, renal parenchymal thickness, however, to date, have not yet been proven. Many of the hemorrhage cases during PCNL could be managed conservatively, however, 0.8% patients required a more invasive procedure to deal with the bleeding14.\n\nTo date there are no specific data available to determine the blood transfusion requirement during PCNL procedures. In our institution, blood units are requested pre-operatively according to clinical estimation. However, this does not always concede to the intra-operative blood loss. This study was aimed at predicting the amount of blood loss during PCNL by identifying the pre-operative factors that could possibly lead to a lower morbidity rate.\n\n\nMaterials and methods\n\nThe present study includes patients who underwent percutaneous nephrolithotomy procedure in our hospital from October 2012 to October 2013. Adult patients (≥18 years old) with pelvic stones >20 mm, inferior calyx stones >10 mm or staghorn stones who agreed to enroll by written informed consent were included in this study. Those with coagulopathy, under anti-coagulant treatment or conversion to open procedure were excluded.\n\nPatients were admitted the day prior to procedure. Stone burden was assessed pre-operatively by multiplying sum of length and width by means of imaging. The PCNL was randomly performed under spinal anesthesia by two endo-urology consultants. The patient was placed in prone position, access gained to pelvicalyceal system with fluoroscopic guidance, followed by dilatation using a metal and fascia dilator before application of sheath. The number and location of punctures were decided intra-operatively, based on pelvicalyceal system. We used pneumatic lithotripsy to break the stone which was subsequently extracted by forceps or grasper. The procedure was completed when the patient was stone-free and any arising complications alleviated.\n\nFull blood counts were taken prior to procedure and thereafter at 12, 24, 36, 72-hours post-operatively. Total blood loss was calculated considering body surface area, sex-adjusted estimated blood volume, initial hematocrit level and 72-hour post-operative hematocrit level. Study protocol has been approved by the Ethical Committee, Faculty of Medicine, Universitas Indonesia (No.89/H2.F1/ETIK/2013).\n\nWe compiled a chart15,16 to assess total blood loss (TBL) to include sex, body surface area, shifting of hematocrit level and amount of blood transfused, as shown in Table 1.\n\nAbbreviations: RBC = red blood cell, BW = body weight\n\nBivariate analysis was done by correlating numerical variables with total blood loss, and associating categorical variables with perioperative blood transfusion amount. Those with significancy of <0.25 were further analyzed with multivariate analysis of linier and logistic regression.\n\n\nResults\n\nA total of 85 PCNL procedures were performed on 85 patients (46 males, 39 females) who completed this study, thus gave statistical power of 0.8. The average age was 50.96 ± 11.87 years. Most of the patients complained of flank pain at initial presentation (Table 2). Staghorn calculi were found in 50.6% patients.\n\nThe mean hematocrit drop was 5.20 ± 3.36%. Average total blood loss was 560.92 ± 428.43 mL with median 511.46 (95% CI: 0.00-1974.84) mL. There were two cases with pelvicalyceal laceration, one with massive hemorrhage (perioperative blood loss 1974.84 mL). There was no significant difference of total blood loss between the cases performed by the two PCNL surgeons (p>0.05).\n\nStepwise multivariate regression analysis which included variables with p-value < 0.25 (Table 3) showed that stone burden was the most influential predictor of blood loss (p=0.037). Meanwhile operative time was found not associated (p-value >0.05) with blood loss. We assessed total blood loss (in mL) as -153.379 + 0.229 × stone burden (mm2) + 0.203 × serum hematocrit baseline (%). This means it is predicted that a kidney stone patient with 1000 mm2 stone burden and baseline hematocrit 40% will have 83.74 mL blood loss during PCNL procedure perioperative.\n\n*Pearson analysis **Spearman analysis\n\nThe average amount of blood units 435.29 ± 114.13 mL was cross-matched for each procedure pre-operatively (Table 4). Nevertheless, 87.1% of patients did not receive blood transfusion perioperatively, thus yielding the blood transfusion rate as 12.9%. Blood transfusion was required by 3.5% patients intra-operatively, 7.1% post-operatively and 2.3% both intra and post-operatively. In total, the cross-matched transfusion ratio was 7.72. The average amount of blood transfused during PCNL procedure: 356.00 ± 145.88 mL.\n\n\nDiscussion\n\nIn this study we did not use conventional, visual estimated blood loss to determine hemorrhage due to high bias factors, subjectivity, persistence of dilution effect and poor accuracy19–23. Laboratory parameters in our study were recorded until 72-hours post-operatively, in order to minimize intravenous hydration and retroperitoneal fluid absorption effects. We used the hematocrit level as the main parameter to determine blood loss rather than hemoglobin to avoid hemodilution effect15,24; Furthermore, the hematocrit level positively correlated with the total blood volume25. It has been reported that also a center in Turkey applies a blood transfusion policy that depends on the hematocrit level (transfusion was indicated when hematocrit level was less than 30%)14.\n\nStone burden was the most influential predictive factor for blood loss during the PCNL procedure in this study, similar to other studies performed. A multivariate analysis showed that complete and partial staghorn calculi were associated with a greater blood loss than with the calyx stones14. Other studies concluded that larger stone burdens11 or staghorn calculi26 required a greater amount of unit blood transfused during the PCNL procedure compared to the smaller stones. Greater hematocrit level changes in staghorn calculi were found during PCNL, while further multivariate tests concluded that staghorn calculi were associated with a greater amount of blood loss (OR 1.92) and a greater decrease in the hemoglobin level compared to non-staghorn cases13. Prolonged and excessive intra-renal maneuver performed for large stone burdens was assumed to increase incidence of injury to renal vasculature11.\n\nOur transfusion rate was similar to the one reported in a retrospective study from Pakistan showing an overall blood transfusion rate of 14.2% with one angioembolization performed to control hemorrhage26. In our study, all cases presenting with massive hemorrhage could be managed conservatively. Lower blood transfusion rates were reported in two other studies from Pakistan, one study from United Kingdom and one study from the United States; these differences occurred due to the younger age group27, supine position used28 and balloon dilatation29,30. A higher transfusion rate (23.8%) compared to our result was shown in a retrospective analysis. An aggressive approach by torqueing the rigid nephroscope to maximize stone clearance at one stage was explained14.\n\nA precise estimation of surgical blood loss is essential in order to avoid excessive usage of blood units. Most of the previous studies emphasize the estimated blood loss rather than the objective prediction. Our calculation of total blood loss included perioperative factors: the patient’s blood volume (based on sex, body weight and height), the number of red cell units transfused, the hematocrit changes, and the amount of hemodilution. To our knowledge, this is the first study that has applied a mathematical approach to predict blood loss during PCNL procedures. We did not include hydronephrosis grading, parenchymal thickness and stone composition to analyze the predictive factors of blood loss, due to the possible limitations relating to our study.\n\n\nConclusions\n\nStone burden was the most influential PCNL blood loss predictive factor in our institution. We estimated that the amount of blood requested and cross-matched for PCNL is much greater than the actual blood loss. Our principle was proposed as a guidance to reduce any unnecessary costs and excessive requirements of blood units.\n\n\nConsent\n\nWritten informed consent to participate in the study and publish clinical data was obtained by the patients.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for Table 3 of 'Blood loss predictive factors and transfusion practice during percutaneous nephrolithotomy of kidney stones', 10.5256/f1000research.8993.d12762931\n\nF1000Research: Dataset 2. Raw data for Table 4 of 'Blood loss predictive factors and transfusion practice during percutaneous nephrolithotomy of kidney stones', 10.5256/f1000research.8993.d12763032",
"appendix": "Author contributions\n\n\n\nFAS - study concepts, design of study, data acquisition, data interpretation, statistical analysis, manuscript preparation. PB - study concepts, design of study, manuscript review, funds collection. NR - design of study, manuscript review, funds collection. FAM - literature research, data acquisition. EN - manuscript editing. MHH - manuscript editing. All authors have agreed to publish this final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nCipto Mangunkusumo Hospital Operational Research Grant 2013 funded this research.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nSaigal CS, Joyce G, Timilsina AR, et al.: Direct and indirect costs of nephrolithiasis in an employed population: opportunity for disease management? Kidney Int. 2005; 68(4): 1808–14. PubMed Abstract | Publisher Full Text\n\nHesse A, Brändle E, Wilbert D, et al.: Study on the prevalence and incidence of urolithiasis in Germany comparing the years 1979 vs. 2000. Eur Urol. 2003; 44(6): 709–13. PubMed Abstract | Publisher Full Text\n\nYasui T, Iguchi M, Suzuki S, et al.: Prevalence and epidemiological characteristics of urolithiasis in Japan: national trends between 1965 and 2005. Urology. 2008; 71(2): 209–13. PubMed Abstract | Publisher Full Text\n\nRomero V, Akpinar H, Assimos DG: Kidney stones: a global picture of prevalence, incidence, and associated risk factors. Rev Urol. 2010; 12(2–3): e86–96. PubMed Abstract | Free Full Text\n\nRahardjo D: Perkembangan penatalaksanaan batu ginjal di RSCM tahun 1997–2002. J I Bedah Indones. 2004; 32: 58–63.\n\nGupta M, Ost M, Shah J, et al.: Percutaneous management of the upper urinary tract. In Wein A, Novick A and Peters C: Campbell's urology. Philadelphia, W.B. Saunders Co, 2007; 1526–1563.\n\nSkolarikos A, Alivizatos G, de la Rosette JJ: Percutaneous nephrolithotomy and its legacy. Eur Urol. 2005; 47(1): 22–8. PubMed Abstract | Publisher Full Text\n\nTürk C, Knoll T, Petrik A, et al.: Guidelines on Urolithiasis.2013; 5: 19–40. Reference Source\n\nPreminger GM, Assimos DG, Lingeman JE, et al.: Chapter 1: AUA guideline on management of staghorn calculi: diagnosis and treatment recommendations. J Urol. 2005; 173(6): 1991–2000. PubMed Abstract | Publisher Full Text\n\nBirowo P, Rasyid N: Sekilas perjalanan hidup dr. Rochani, Sp.B, Sp.U(K). In Birowo P: Purna bakti dr. Rochani, Sp.B, Sp.U(K). Jakarta, Divisi Urologi Departemen Ilmu Bedah FKUI/Departemen Urologi RSCM, 2010.\n\nKukreja R, Desai M, Patel S, et al.: Factors affecting blood loss during percutaneous nephrolithotomy: prospective study. J Endourol. 2004; 18(8): 715–22. PubMed Abstract | Publisher Full Text\n\nKeoghane SR, Cetti RJ, Rogers AE, et al.: Blood transfusion, embolisation and nephrectomy after percutaneous nephrolithotomy (PCNL). BJU Int. 2013; 111(4): 628–32. PubMed Abstract | Publisher Full Text\n\nAkman T, Binbay M, Sari E, et al.: Factors affecting bleeding during percutaneous nephrolithotomy: single surgeon experience. J Endourol. 2011; 25(2): 327–33. PubMed Abstract | Publisher Full Text\n\nTurna B, Nazli O, Demiryoguran S, et al.: Percutaneous nephrolithotomy: variables that influence hemorrhage. Urology. 2007; 69(4): 603–7. PubMed Abstract | Publisher Full Text\n\nRosencher N, Kerkkamp HE, Macheras G, et al.: Orthopedic Surgery Transfusion Hemoglobin European Overview (OSTHEO) study: blood management in elective knee and hip arthroplasty in Europe. Transfusion. 2003; 43(4): 459–69. PubMed Abstract | Publisher Full Text\n\nHurle R, Poma R, Maffezzini M, et al.: A simple mathematical approach to calculate blood loss in radical prostatectomy. Urol Int. 2004; 72(2): 135–9. PubMed Abstract | Publisher Full Text\n\nKhan FA, Khan M, Ali A, et al.: Estimation of blood loss during Caesarean section: an audit. J Pak Med Assoc. 2006; 56(12): 572–5. PubMed Abstract\n\nSeruya M, Oh AK, Rogers GF, et al.: Blood loss estimation during fronto-orbital advancement: implications for blood transfusion practice and hospital length of stay. J Craniofac Surg. 2012; 23(5): 1314–7. PubMed Abstract | Publisher Full Text\n\nPoletajew S, Antoniewicz AA: Blood loss during laparoscopic radical prostatectomy - is it significant or not? Cent European J Urol. 2012; 65(1): 11–13. PubMed Abstract | Publisher Full Text\n\nMcCullough TC, Roth JV, Ginsberg PC, et al.: Estimated blood loss underestimates calculated blood loss during radical retropubic prostatectomy. Urol Int. 2004; 72(1): 13–16. PubMed Abstract | Publisher Full Text\n\nStafford I, Dildy GA, Clark SL, et al.: Visually estimated and calculated blood loss in vaginal and cesarean delivery. Am J Obstet Gynecol. 2008; 199(5): 519.e1–7. PubMed Abstract | Publisher Full Text\n\nGharoro EP, Enabudoso EJ: Relationship between visually estimated blood loss at delivery and postpartum change in haematocrit. J Obstet Gynaecol. 2009; 29(6): 517–520. PubMed Abstract | Publisher Full Text\n\nSchorn MN: Measurement of blood loss: review of the literature. J Midwifery Womens Health. 2010; 55(1): 20–27. PubMed Abstract | Publisher Full Text\n\nKukreja R, Desai M, Patel S, et al.: Factors affecting blood loss during percutaneous nephrolithotomy: prospective study. J Endourol. 2004; 18(8): 715–722. PubMed Abstract | Publisher Full Text\n\nTurna B, Nazli O, Demiryoguran S, et al.: Percutaneous nephrolithotomy: variables that influence hemorrhage. Urology. 2007; 69(4): 603–607. PubMed Abstract | Publisher Full Text\n\nZehri AA, Biyabani SR, Siddiqui KM, et al.: Triggers of blood transfusion in percutaneous nephrolithotomy. J Coll Physicians Surg Pak. 2011; 21(3): 138–41. PubMed Abstract\n\nMahmud M, Zaidi Z: Percutaneous nephrolithotomy in children before school age: experience of a Pakistani centre. BJU Int. 2004; 94(9): 1352–4. PubMed Abstract | Publisher Full Text\n\nRana AM, Bhojwani JP, Junejo NN, et al.: Tubeless PCNL with patient in supine position: procedure for all seasons?--with comprehensive technique. Urology. 2008; 71(4): 581–5. PubMed Abstract | Publisher Full Text\n\nArmitage JN, Irving SO, Burgess NA, et al.: Percutaneous nephrolithotomy in the United kingdom: results of a prospective data registry. Eur Urol. 2012; 61(6): 1188–93. PubMed Abstract | Publisher Full Text\n\nTomaszewski JJ, Smaldone MC, Schuster T, et al.: Factors affecting blood loss during percutaneous nephrolithotomy using balloon dilation in a large contemporary series. J Endourol. 2010; 24(2): 207–11. PubMed Abstract | Publisher Full Text\n\nSyahputra FA, Birowo P, Rasyid N, et al.: Dataset 1 in: Blood loss predictive factors and transfusion practice during percutaneous nephrolithotomy of kidney stones: a prospective study. F1000Research. 2016. Data Source\n\nSyahputra FA, Birowo P, Rasyid N, et al.: Dataset 2 in: Blood loss predictive factors and transfusion practice during percutaneous nephrolithotomy of kidney stones: a prospective study. F1000Research. 2016. Data Source"
}
|
[
{
"id": "14721",
"date": "30 Jun 2016",
"name": "M. Hammad Ather",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors have assessed the predictive value of various parameters for transfusion during / following PCNL. It is an important area as bleeding and need for transfusion is still a major concern during PCNL. I read with interest the fact that the authors have performed all procedures under spinal anesthesia. Although it is feasible and has been reported in literature it is both cumbersome for patients, anesthetist and surgeon to perform this procedure in an awake patient. In addition to maintain level of spinal anesthesia to provide pain free procedure and yet not compromising ventilation in a spontaneously breathing patient is a difficult task. It is not clear from the methods described if the procedure was performed in prone or supine position. In the methods section authors have noted that \"the procedure was randomly performed.....\", I am not sure as to what the authors intended to mean. In addition in the methods section authors have noted that the \"The procedure was completed when the patient was stone-free and any arising complications alleviated. \" the meaning of this sentence is also not clear. In the results section authors have noted that \"statistical power of 0.8 \", kindly clarify the sentence.\nAuthors noted stone burden as clinical (statistically significant) factor in predicting need for transfusion. This is conformed by many previous studies and authors conclusion does not surprise me.\nIn recent years many scoring systems have been developed to predict complexity of the procedure and stone free rate like STONE nephrolithomeetry score, Guys score etc. Authors should comment why they decided to use stone burden rather than these scores in predicting complexity level of the procedure and predicting need for transfusion.\nA recent report by Un et al., 20151 noted transfusion requirement of ~10% and need for embolization in <1% in over 1400 PCNL over a 7 years period.",
"responses": []
},
{
"id": "14722",
"date": "11 Jul 2016",
"name": "Manint Usawachintachit",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a study looking for predictive factor of blood loss following PCNL from a single institution for 85 patients over a 1-year period. It’s interesting data in that while most relevant published studies were carried out retrospectively, this study is prospective. The primary endpoints were changes in hematocrit level and blood transfusion requirement over an early postoperative period.\n\nOverall, the paper was written concisely with a good methodology to track blood loss level. I would recommend some revision enumerated below:\n\nGeneral comments\nBesides grammatical errors, there are some misspelled words.\n\nAbstract\nIn the Result, a percentage of intra and postoperative transfusion should be 2.3%, not 23%. It may be helpful to add a conclusion to the abstract.\n\nIntroduction\nNo specific comments.\n\nMaterial and Methods\nWere all patients included in this study consecutively? How was stone burden assessed by means of imaging? Was it measured from CT-scan, plain film, or IVP? How was a hematocrit level measured? Was it from a full blood draw or just fingertip needle sticks crude measurement?\n\nWhat was an algorithm of decision for perioperative blood transfusion in this study? Did the authors make a judgement based on a “specific” cutoff point of the most recent blood count (at 12, 24, 36, and 72 hours after PCNL) or make a decision subjectively? There should be a systematic methodology in this issue because it significantly affected transfusion rate and volume.\n\nResults\nHow many patients were excluded from this study because of a conversion to open procedure? Were they converted from profound intraoperative bleeding? What were a mean BMI of patients, and distribution of hydronephrosis in this series? The authors have mentioned in the Introduction that they were risk factors for blood transfusion flowing PCNL. What was a distribution of renal access (upper/middle/lower/multi-tract) in this study? Did number of renal access tract affect blood loss or transfusion rate? In a multivariate analysis of variables and blood loss, was stone burden (p = 0.037) the only one factor that remained statistically significant? This data was not shown in Table 3. What were other variables included in the multivariate analysis and why did the authors include operative time (p > 0.05) in this model?\n\nDiscussion\nThe authors mentioned a hematocrit cut-point of 30% from a Turkish study. Did the authors use the same cut-point for transfusion in this study? Transfusion rate in this series was relatively high (12.9%) and the authors have demonstrated a significant correlation between stone burden and blood loss. Could this high transfusion rate cause by a larger stone burden? This data (mean stone burden) was not shown in the manuscript. What is a clinical benefit of the mathematical equation to predict blood loss preoperatively? Could we use it to obviate blood request in some patients?\n\nMiscellaneous\nIn Table 2, it may be more understandable to order presenting symptoms by percentages. In Dataset 1, why did the authors perform PCNL in some stones of 5 mm2 size. Is it more reasonable to treat them with ESWL or flexible ureteroscopy.",
"responses": []
},
{
"id": "14725",
"date": "12 Jul 2016",
"name": "Doddy Musbadianto Soebadi",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a nice article concerning blood loss during percutaneous nephrolithotomy (PCNL) procedures. We need these data to confirm that this procedure is safe to be done in a proper way.\nThe authors have meticulously analyzed and discussed the results.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1550
|
https://f1000research.com/articles/5-1549/v1
|
30 Jun 16
|
{
"type": "Review",
"title": "Circadian clock regulation of skeletal muscle growth and repair",
"authors": [
"Somik Chatterjee",
"Ke Ma",
"Somik Chatterjee"
],
"abstract": "Accumulating evidence indicates that the circadian clock, a transcriptional/translational feedback circuit that generates ~24-hour oscillations in behavior and physiology, is a key temporal regulatory mechanism involved in many important aspects of muscle physiology. Given the clock as an evolutionarily-conserved time-keeping mechanism that synchronizes internal physiology to environmental cues, locomotor activities initiated by skeletal muscle enable entrainment to the light-dark cycles on earth, thus ensuring organismal survival and fitness. Despite the current understanding of the role of molecular clock in preventing age-related sarcopenia, investigations into the underlying molecular pathways that transmit clock signals to the maintenance of skeletal muscle growth and function are only emerging. In the current review, the importance of the muscle clock in maintaining muscle mass during development, repair and aging, together with its contribution to muscle metabolism, will be discussed. Based on our current understandings of how tissue-intrinsic muscle clock functions in the key aspects muscle physiology, interventions targeting the myogenic-modulatory activities of the clock circuit may offer new avenues for prevention and treatment of muscular diseases. Studies of mechanisms underlying circadian clock function and regulation in skeletal muscle warrant continued efforts.",
"keywords": [
"Suprachiasmatic nuclei",
"Clock Controlled Genes",
"Myogenic Progenitor Cells",
"Myosin Heavy Chain"
],
"content": "Abbreviations\n\nSCN: Suprachiasmatic nuclei\n\nCCGs: Clock Controlled Genes\n\nMPCs: Myogenic Progenitor Cells\n\nMyHc: Myosin Heavy Chain\n\n\nIntroduction\n\nThe circadian clock is the overt ~24-hour daily rhythm in physiology and behavior that evolved to respond to earth’s rotation. This evolutionarily-conserved mechanism synchronizes diverse internal biological processes with environmental timing cues to ensure organismal adaptation, fitness and survival1–3. The circadian clock system consists of a hierarchal organization. The central clock resides in the suprachiasmatic nuclei (SCN) of the hypothalamus and transmits timing signals from light inputs to drive peripheral tissue clocks1–3. Nearly all tissue/cell types in the body possess cell-autonomous clock circuits that are entrained by central clock signals, but can be fully uncoupled through diet timing manipulations such as restricted feeding1,3–5. In recent years, the clock system in skeletal muscle has been recognized to play critical roles in key aspects of skeletal muscle physiology ranging from structural maintenance to functional regulation6–9. As locomotor activity, the essential function of skeletal muscle in all animal species is under direct circadian clock control through sleep-wake cycles, and the intimate interplay between clock and skeletal muscle physiology is evolutionarily-conserved to ensure fitness and survival. It is therefore possible that the current understandings of the intricate interactions between circadian clock regulation and skeletal muscle at transcriptional, functional and organismal levels are merely at the beginning stages of our endeavor.\n\n\nThe tissue-intrinsic circadian clock in skeletal muscle\n\nMost physiological processes and diurnal activities of organisms follow distinct daily oscillations, governed via environmental cues by the circadian time-keeping system. This hierarchal machinery is composed of a central pacemaker in the brain’s SCN and peripheral clocks in nearly every tissue and cell types, driven by the central clock pacemaker under normal physiological conditions. The complex interplays between central and peripheral clock systems function in concert to exert proper temporal control on various circadian physiological outputs. At the molecular level, an intricate transcriptional-translational network of circadian clock circuit that generates circadian rhythmicity has been well-defined, although novel modulators of the circadian clock loop continue to emerge. The positive and negative regulators of the molecular clock network are reciprocally regulated through intricate transcriptional and translational feedback loops10. Bmal1 (Brain and Muscle Arnt-like 1) and CLOCK (Circadian Locomotor Output Cycles Kaput), two transcription activators of the molecular clock, form a heterodimer that turns on transcription of its negative regulators. These regulators, Pers (Period1, 2 and 3), Crys (Cryptochrome1 and 2), bind to CLOCK-Bmal1 and inhibit transcriptional activation; whereas the Rev-erbs (Rev-erbα and Rev-erbβ) are direct transcriptional repressors of Bmal1. Notably, Bmal1, the essential transcriptional activator of the molecular clock, is highly expressed in skeletal muscle and initiates target genes transcription through binding canonical E-box, or E’ sequences11. The transcriptional repressors, Rev-erbs, bind to RORE sequence and together with the activator RORα (RAR-related orphan receptor α), generate the circadian oscillatory control of Bmal1 expression. ChIP-sequencing studies in liver have demonstrated extensive overlap of genome-wide cis-acting target promoter sequences between Bmal1 and Rev-erbα/Rev-erbβ12,13. This suggests that the components of the molecular clock network function coordinately to generate the circadian rhythmicity of their target genes in peripheral tissues, including skeletal muscle. Interestingly, although embryonic stem cells express clock genes, they do not display overt circadian rhythmicity14. The gradual acquisition of diurnal oscillation in clock genes, such as Bmal1 and DBP (D site of albumin promoter binding protein), accompanies their cellular differentiation. This observation raises an intriguing notion of possible coupling between cellular developmental processes with the acquirement of molecular circadian rhythms. Skeletal muscle, the most abundant tissue in mammals that dictates physical activity, possess self-sustaining endogenous molecular clock15.\n\nThe circadian clock network plays a prominent role in maintenance of skeletal muscle mass, with the loss of Bmal1 leading to severe sarcopenia with age16. Numerous studies to date involving animal models harboring specific clock gene deletions or mutations have provided useful genetic tools to dissect the roles of the clock circuit in skeletal muscle, as summarized in Table 1. These studies provide strong evidence attesting the importance of circadian clock functions in modulating various aspects of skeletal muscle physiology, including muscle growth and maintenance, contractile performance, structural organization, glucose metabolism and energy production. A remarkable 17% of genes exhibit circadian-like oscillations in skeletal muscle, and nearly 30% of those circadian transcripts lose their rhythmicity in CLOCK-mutant mice17. This indicates that the molecular clock plays a central role in conferring appropriate temporal regulation of clock-controlled genes (CCGs) in skeletal muscle. MyoD1 (Myogenic Differentiation 1), a key transcriptional regulator activated during the early stages of myoblast differentiation and muscle development, has emerged as a CCG based on its distinct circadian expression pattern in adult muscle6. Ablation of core components of the molecular clock, CLOCK or Bmal1, blunts MyoD1 circadian expression as well as its target genes, which is associated with disruption of myofiber sarcomeric organization and muscle contractile function6,17. Findings of similarly impaired functional deficits in muscle specific force generation in the CLOCK mutant and Bmal1-deficient mice indicate the concerted clock contribution to this essential skeletal muscle function6,16,18,19. Furthermore, Per1, Per2, as well as ROR-deficient mice were found to exhibit related pathologies in muscle structure and function, such as muscle weakness, contractile and locomotor deficits6,20–22, further supporting the notion that the clock function is required for skeletal muscle activities. However, so far, the mechanistic link between clock and muscle functional regulation has not been clearly defined.\n\nAccumulating evidence indicates an intimate interplay between circadian clock machinery and metabolic regulations, either at the level of temporal control evident in many key metabolic processes in distinct metabolic tissues, or in the maintenance of whole-body metabolic homeostasis8,17,23–25. In skeletal muscle, a key organ for metabolic substrate oxidation, nearly 30% of CLOCK-differentially regulated transcripts are involved in metabolism17. Both Bmal1 and Rev-erbα deficiency in mice alters mitochondrial morphology, content or oxidative function6,9.\n\nSo far however, as the majorities of studies of clock function in skeletal muscle are confined to the use to whole-body global ablation models, central clock contribution or secondary effects from other tissues may confound certain findings. Future studies are required to interrogate functions of the intrinsic muscle clock independently of central clock regulation imparted on muscle function. In addition, specific temporal controls conferred by the intrinsic muscle clock may differ in distinct cell types and may be specific to developmental stages. Therefore, there is an urgent need to critically assess the full-range of roles of the intrinsic muscle clock in muscle through developmental stage-selective and tissue or muscle cell type-restricted genetic models.\n\n\nClock modulation of muscle growth, repair and mass maintenance\n\nThe first indication that the clock is involved in skeletal muscle maintenance comes from the dramatic phenotype of aging-associated sarcopenia found in Bmal1-null mice16. At 40-weeks of age, the genetic loss of Bmal1 led to a reduction of nearly half of the normal muscle weight with dramatically shortened life span, suggesting a premature aging phenotype in these mice. Interestingly, the lower muscle mass manifests as early as in 8-weeks old mice, when satellite cells are at the peak of their proliferative capacity7. The maintenance of muscle mass encompasses two distinct contributions, one involving myonuclear accretion due to myogenic progenitor proliferation and maturation in early postnatal growth, and mature myofiber hypertrophy in adult stage26. Thus, these findings collectively suggest that the marked reduction of muscle weight in adult Bmal1-null mice may result from the combination of a developmental defect and impaired hypertrophic growth. Furthermore, specific rescue of Bmal1 expression in skeletal muscle was able to prolong survival of Bmal1-null animals, whereas brain-specific rescue was not sufficient19, highlighting that the muscle-intrinsic clock is critical for maintaining proper ambulatory activity that is essential for survival. Miller et al. have demonstrated that Bmal1 is required for various aspects essential for proper muscle performance including sarcomeric structure, mitochondrial morphology and muscle contractile activities6, which could be the structural and functional impairments underlying the severe premature aging-like muscle defects observed in Bmal1-deficient animals. Further detailed investigations into the molecular pathways mediating these profound clock effects in skeletal muscle are warranted, particularly in the absence of central clock dysfunction. An intriguing finding is the substantial similarity observed between the sarcomeric disorganization of the Bmal1-null and the CLOCK mutant mice with that of the MyoD-null mutants6. The underlying mechanism linking clock with muscle structure/function regulation could be attributed to the direct transcriptional activation of the Bmal1/CLOCK complex of the identified MyoD1 enhancer element, although non-consensus E-box sequences are involved6,27. In vivo, enhanced expression of the myogenic regulatory factors MyoD1 and myogenin was detected during dark hours, although this diurnal rhythm is strongly suppressed by fasting28. During embryonic development, MyoD1, together with Myf5, specifies the myotome and drives myogenesis29,30. Thus, this identified specific link of molecular clock with MyoD1 transcription raises an intriguing question as to whether the muscle intrinsic clock participates in muscle development processes or myogenesis. Remarkably, on a genome-wide scale, surveying of CLOCK-controlled mRNA expression in the skeletal muscle reveals that growth, proliferation and differentiation processes comprise a significant 15% of the overall transcripts6. In agreement with this finding, work from our group demonstrated that Bmal1 is a key positive regulator to promote myogenic differentiation7, and its regulation of proliferative behavior and expansion of myogenic progenitor cells is required for tissue regeneration upon injury8. As an evolutionarily-conserved machinery to anticipate and adapt to environmental cues, circadian clock has been implicated in transcriptional control of developmental signaling pathways important for stem cell modulation during tissue remodeling processes31–33. The clock may provide critical temporal cues to orchestrate the highly ordered stem cell activation, proliferation and differentiation processes required for tissue development, physiological turnover or regenerative repair. The distinct developmental signals required for tissue homeostasis may reflect its specific developmental and functional needs. In skeletal muscle, we found Bmal1 exerts circadian time-dependent transcriptional control on key components of the canonical Wnt signaling pathway34. When tested in muscle injury-elicited regeneration models, including cardiotoxin-induced and freezing injury, mice lacking Bmal1 displayed a significant defect in regenerative myogenic response accompanied by attenuated repair8. Furthermore, the satellite cell expansion process, a major component to ensure proper regeneration, is also impaired due to reduced proliferative capacity. This is likely attributed to Bmal1 regulation of Wnt signaling, since loss of Bmal1 leads to blunted Wnt signaling as observed in Bmal1-null mice muscle regeneration6–8. Wnt signaling drives embryonic development of the skeletal muscle lineage34, and plays important roles in modulating adult muscle satellite cell functions35,36. Our original findings provide strong evidence for the cell-autonomous roles of molecular clock in myogenic progenitor cells (MPCs), which provide the major cellular source for muscle growth and repair. This mechanism likely mediates, at least in part, the demonstrated importance of clock function in muscle mass maintenance, particularly during the early postnatal development. Muscle homeostasis and remodeling requires contribution from muscle satellite cells, and their proliferative capacity declines with age. Thus, whether the sarcopenia observed in Bmal1-null mice that resemble early aging could be mediated at least in part by declining clock function in the muscle warrants further investigation. In contrast, as satellite cells are largely not required or necessary for adult skeletal muscle hypertrophic growth37, another possibility is that clock may function in hypertrophic signaling pathways in mature myofibers to contribute to adult muscle mass regulation. These questions could be addressed by muscle developmental-stage specific animal models using currently available genetic tools.\n\nAlthough the major body of research to date has been focused on the role of Bmal1 as a clock transcription activator, cytosolic Bmal1 was recently identified as a factor facilitating protein translation that links the circadian network and the mTOR (Mechanistic Target of Rapamycin) signaling pathway38. Most intriguingly, the Bmal1-mediated mTOR circadian modulation of translation activities is controlled by daily oscillatory magnesium levels in cells39. These recent findings raise the possibility that Bmal1 and the clock could directly participate in muscle hypertrophic pathways via post-transcriptional mechanisms. mTOR signaling, activated by upstream growth factors and PI3 kinase-Akt phosphorylation, is a major regulatory mechanism that promotes protein synthesis to induce skeletal muscle hypertrophy26,40. In addition, PI3K-Akt-mTOR signaling suppresses muscle atrophy40,41. Interestingly, multiple components of the Akt/mTOR signaling pathway are reported to be under circadian regulation. Circadian patterns of expression were detected for Akt1 and ribosomal protein S6 of the hypertrophic signaling, and MuRF1 and Fbxo32 within the atrophic pathway in skeletal muscle28. Notably, the circadian profile of Akt1 phosphorylation, an indicator of in vivo activity, persists at fasting despite lower levels than ad-libitum feeding, indicating an endogenous rhythm independent of food signals. However, as feeding cycle is dominant zeitgeber for peripheral clocks such as the muscle, there are strong interplays between circadian oscillatory patterns and feeding-fasting switch.\n\nThe skeletal muscle phenotypes found in genetic models of additional clock genes further support the notion that the molecular clock as a regulatory circuit exerts profound influence on skeletal muscle mass and function. Both the clock repressor, Rev-erbα, and its reciprocal transcription activator RORα on the RORE responsive element have been implicated in the regulation of myogenic differentiation42,43. Whereas the constitutive expression of dominant negative Rev-erbα promotes myogenic progression42,44, myogenic differentiation and myogenic pathways gene expression are suppressed by muscle-specific expression of a truncated RORα mutant43. Importantly, the loss of Rev-erbα deficient mice was found to display lower body weight and altered myosin heavy chain (MyHC) isoform expression with a fast-to-slow MyHC isoform transformation in skeletal muscle, suggesting its involvement in muscle mass maintenance and metabolic control45. The findings of opposing actions of Rev-erbα vs. RORα on myogenic pathways, as well as the opposite effects of clock repressor Rev-erbα vs. activator Bmal1 on myogenesis, strongly suggest orchestration of circadian clock gene functions in regulation of myogenic precursor development. Currently, the molecular mechanisms mediating Rev-erbα vs. RORα actions on myogenesis has not been addressed. Furthermore, based on the significantly increased muscle mass demonstrated in the mPer2-null mice, a potential negative effect of the Bmal1/CLOCK inhibitory regulator, Period 2 (Per2), on muscle growth has been suggested21. Per2 functions in the myogenic cascade remain to be seen. Surprisingly, mPer2 and mPer1 functions in the skeletal muscle are distinct, as the altered muscle mass and metabolic pathways are only evident in the mPer2-null mice but not mPer1-deficient animals. Another transcription inhibitor of CLOCK/Bmal1 function, the basic helix-loop-helix factor Dec2/Sharp1, can suppress myogenic differentiation through its inhibitory interaction with MyoD46,47.\n\nTaken together, studies of mice harboring genetic mutations of clock genes to date have clearly established a strong link between the molecular clock circuit as a whole and maintenance of skeletal muscle development, growth and potentially hypertrophy. Further studies will be needed to address whether other types of clock disruptions, such as those induced by the dys-synchrony between environmental lighting cycle with endogenous circadian clock cycles, may influence muscle growth and remodeling process. In our post-industrial society, the so-called “social jetlag”, referring to the discordance between our activity/sleep cycle vs. clock cycles, may contribute to the development of certain type of muscle diseases, particularly in the aging population with frequent sleep disorders48. The concerted regulatory functions of the muscle intrinsic clock machinery in maintaining skeletal muscle mass may be important mechanisms to protect against muscle loss in aging-associated or chronic disease-induced muscle wasting conditions. Investigations of underlying molecular pathways mediating clock function in muscle may, therefore, reveal novel therapeutic targets for muscle disease treatment.\n\n\nClock regulation of skeletal muscle structure and function\n\nA major output of circadian clock in animals is its tight control of locomotor activity cycles. As an evolutionarily-conserved mechanism that enables entrainment to the light-dark cycles on earth, the strict behavioral circadian rhythmicity of animals ensures their survival and fitness. Thus, it is not surprising that it has long been recognized that in humans, skeletal muscle torque, strength and power are higher in the late afternoon, between 16:00 and 18:00 hours than compared to the morning49–52,53. Major indexes of athletic performance abilities, such as muscle strength, reaction time and flexibilities, display significant time-of-the day dependence54,55. Knee extensor muscles exhibit a typical diurnal pattern in maximal isometric strength measured in male athletes, which peaks at mid-to-late afternoon period (16:00–20:00 hours)56. Interestingly, partial sleep deprivation was found to have a detrimental effect on the power output of muscle performances, although this effect may depend on the time of the day of the measurements or the onset timing and duration of the sleep disruption57,58. These findings suggest potentially intimate interplay between clock control, either central or muscle-intrinsic, and physical activity. Most importantly, under various experimental settings, increase in activity level, such as exercise, has been shown to entrain core clock genes and CCGs in humans59 as well as in equine skeletal muscle60. Resistance exercise is capable of shifting expression of diurnally-regulated genes in human skeletal muscle by inducing genes that are normally repressed, while down-regulating genes that are highly expressed59. On the other hand, loss of muscle activity by unilateral sciatic nerve denervation leads to marked atrophy, and reduces the expression of many core clock genes, including Bmal1, Per1, RORα and Rev-erbα in mouse skeletal muscle61. Notably, activity cycles can impact the central clock rhythm. Restricted wheel access in mice, which enforces inverse activity cycles, significantly delays re-entrainment to normal light/dark rhythm62. Together, these studies suggest that physical activity in animals could function as a strong clock entrainment signal, particularly for the skeletal muscle clock. Thus a potential feedback regulatory relationship exists between the circadian clock network and muscle function.\n\nThe skeletal muscle circadian transcriptome was first reported by Miller et al., based on analysis of gene expression from muscle collected every 4 hours over two circadian cycles17. In skeletal muscle, proteins involved in the regulation of gene transcription are abundant, representing ~17% of rhythmic genes in muscle17. This indicates that many essential functions and physiological processes in skeletal muscle are influenced by the transcriptional output of the clock. Interestingly, a high proportion of cycling transcripts peak midway through the dark phase in mice, coinciding with the peak period of physical activity and feeding in nocturnal species. Particularly, a single large cluster of rhythmic genes displays peak expression at Circadian Time 18 (CT18) of the midpoint of the active phase for mice, even under constant darkness17. However, how much of these processes require central or skeletal muscle-specific molecular clock function has not yet been fully established. Based on the observation that resistance exercise can directly affect expression of key clock components and downstream targets in human skeletal muscle59, the peak expression of rhythmic transcripts in muscle could be attributed to the orchestration of the endogenous muscle clock control and central clock-induced locomotor activity rhythm. Interestingly, although repeated exercise can induce phase-shift of the clock in skeletal muscle, the SCN rhythms are not affected15. Thus, locomotor activity may phase-coordinate the intrinsic rhythmic expression of genes in skeletal muscle with central clock-controlled sleep/wake cycles under normal physiological conditions. These findings together indicate intimate interplays between muscle physical activity and the molecular clock machinery in skeletal muscle, although the underlying mechanistic links, particularly how activity-stimulated signals in muscle is transmitted to clock resetting, phase or amplitude modulation, remain to be elucidated.\n\n\nClock participation in muscle metabolism\n\nThe molecular clock machinery governs the temporal control in metabolic processes24. Disruption of this regulatory mechanism profoundly altered metabolic homeostasis leading to the development of obesity and insulin resistance63–67. Skeletal muscle comprises approximately 40% of the body mass of most mammals, and functions as a major site for glucose disposal and lipid oxidation. Skeletal muscles account for approximately 85% of postprandial insulin-mediated glucose disposal, and changes in muscle function contribute to insulin resistance and metabolic syndromes68. Thus, given its prominent role in temporal control of metabolism, the cell-intrinsic clock machinery in skeletal muscle could be critical for whole-body metabolic homeostasis. There is increasing interest in understanding how the endogenous circadian clock functions to modulate muscle metabolism.\n\nThe role of the endogenous skeletal muscle molecular clock in regulating muscle metabolic functions and whole body metabolic homeostasis has emerged recently17,69,70. Initial studies of differentially-regulated genes in CLOCK mutants studies indicate that a remarkable ~35% percentage of rhythmic genes in muscle are involved in metabolism17. Further, analysis of circadian metabolic genes revealed a temporal separation of genes involved in substrate utilization vs. storage over a daily period, suggesting a clock-controlled orchestration of distinct catabolic and anabolic metabolic pathways in skeletal muscle70.\n\nTo address the contribution of skeletal muscle to whole body circadian energy homeostasis, skeletal muscle-specific Bmal1 deletion was created to test the function of Bmal1 in skeletal muscle glucose metabolism69,70. Muscle-specific deletions of Bmal1, either constitutively or through inducible-Cre lines, cause impaired insulin-dependent glucose uptake and reduced glucose oxidation in skeletal muscle69. While canonical insulin signaling pathway is not affected, the level of GLUT4 glucose transporter responsible for glucose uptake was significantly lower. It is interesting that these defects in glucose utilization do not lead to overt changes in insulin sensitivity, possibly due to compensatory mechanisms in other tissues. Applying a global gene expression profiling approach in an inducible mouse model of Bmal1 ablation in muscle, a later study revealed significantly altered expression of genes involved in metabolic substrate oxidation70. Significant down-regulation of circadian genes involved in glucose utilization were observed, along with significant up-regulation of genes involved in lipid metabolism. This gene expression profile suggests muscle fiber type switch to a slow oxidative fiber-type consistent with a substrate shift from carbohydrate to lipid utilization, although the precise fiber type distribution in fast or slow muscle fibers were not assessed70. Thus, two independent studies suggest that the endogenous molecular clock may coordinate skeletal muscle metabolic substrate utilization with metabolite availability occurring during fasting-feeding transitions balance, which could play a significant role in whole-body energy partitioning between tissues to maintain metabolic homeostasis10.\n\nThe circadian clock repressor gene, Rev-erbα, is known to play important roles in metabolic regulations71–73. In skeletal muscle, Rev-erbα was found to be highly expressed in oxidative fiber types, and promotes skeletal muscle oxidative capacity through inhibition of mitochondria autophagy and abundance9. A previous study indicated that there was significant fast-to-slow MyHC isoform transformation in Rev-erbα-deficient mice, albeit only in soleus muscle45. Most importantly, as a ligand-dependent nuclear receptor, Rev-erbα is amenable to synthetic ligand modulations. Synthetic agonists of Rev-erbα, display potent anti-obesity and lipid lowering efficacy in mice74. Notably, the activation of Rev-erbα by synthetic agonists induces fatty acid oxidation pathways while suppresses lipid synthesis genes in skeletal muscle, likely a significant contributor to its lipid-lowering effects in vivo. In contrast, the exercise endurance of Rev-erbα-deficient mice is reduced, likely a result of lower mitochondrial function in muscle; whereas the activation of Rev-erbα by an agonist improves endurance capacity9. Additional studies of Rev-erbα-deficiency on metabolic homeostasis reveal mild hyperglycemia and increased fatty acid utilization, indicating that Rev-erbα may promote the preferential use of glucose at the expense of peripheral lipid utilization73. These studies establish a foundation to further explore the mechanistic basis of Rev-erbα as a “druggable” target for metabolic diseases, and the potential of modulating the tissue clock circuit as therapeutic strategies. On the other hand, in muscle cells, the dominant negative mutant of RORα, the transcriptional activator of RORE-harboring promoters antagonistic to Rev-erbα, inhibits expression of many genes involved in lipid homeostasis, including carnitine palmitoyltransferase-1 for fatty acid oxidation75. Given that the global loss in the staggerer mice leads to reduced muscle strength and hypo-α-lipoproteinemia76, the in vivo effects of RORα inhibition in muscle metabolism remains to be seen. In line with findings of the molecular clock regulation in glucose metabolic homeostasis, the loss of Cry1 and Cry2 in mice induces systemic glucose intolerance, although whether this defect is a result of altered muscle glucose disposal needs further detailed studies77.\n\nTaken together, current findings indicate that the clock machinery in skeletal muscle plays a significant role in orchestrating metabolic substrate metabolism. As feeding signals are strong clock entrainment cues, whether clock functions as a temporal mechanism to adapt to feeding-fasting induced metabolic substrate switching remains to be studied. Future investigation into the molecular mechanisms linking clock and muscle metabolic substrate flux may yield novel targets for disease treatment including obesity and diabetes.\n\n\nConclusion\n\nThe circadian clock plays key roles in critical aspects of skeletal muscle physiology. Thus, it is imperative to dissect the precise underlying mechanisms involved in these multifaceted interactions. Studies of the intimate interplays of the tissue-intrinsic clock with growth, hypertrophy, activity and metabolism in skeletal muscle would provide a wealth of novel targets for disease prevention or treatment. Particularly, given the importance of the circadian clock network in muscle mass maintenance, interventions targeting myogenic-modulatory activities of the clock circuit may offer new avenues for the prevention and treatment of muscular diseases, particularly those associated with circadian dysregulation.\n\n\nAbbreviations\n\nSCN: Suprachiasmatic nuclei\n\nCCGs: Clock Controlled Genes\n\nMPCs: Myogenic Progenitor Cells\n\nMyHc: Myosin Heavy Chain",
"appendix": "Author contributions\n\n\n\nBoth SC and KM conceived and wrote the article, and approved the final version.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis project is supported by grants from the American Heart Association 12SDG12080076, American Diabetes Association 1-13-BS-118, and Muscular Dystrophy Association 381294 to K. M.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nDibner C, Schibler U, Albrecht U: The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu Rev Physiol. 2010; 72: 517–549. PubMed Abstract | Publisher Full Text\n\nReppert SM, Weaver DR: Coordination of circadian timing in mammals. Nature. 2002; 418(6901): 935–941. PubMed Abstract | Publisher Full Text\n\nSchibler U, Sassone-Corsi P: A web of circadian pacemakers. Cell. 2002; 111(7): 919–922. PubMed Abstract | Publisher Full Text\n\nDamiola F, Le Minh N, Preitner N, et al.: Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev. 2000; 14(23): 2950–2961. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLe Minh N, Damiola F, Tronche F, et al.: Glucocorticoid hormones inhibit food-induced phase-shifting of peripheral circadian oscillators. EMBO J. 2001; 20(24): 7128–7136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndrews JL, Zhang X, McCarthy JJ, et al.: CLOCK and BMAL1 regulate MyoD and are necessary for maintenance of skeletal muscle phenotype and function. Proc Natl Acad Sci U S A. 2010; 107(44): 19090–19095. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatterjee S, Nam D, Guo B, et al.: Brain and muscle Arnt-like 1 is a key regulator of myogenesis. J Cell Sci. 2013; 126(Pt 10): 2213–2224. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatterjee S, Yin H, Nam D, et al.: Brain and muscle Arnt-like 1 promotes skeletal muscle regeneration through satellite cell expansion. Exp Cell Res. 2015; 331(1): 200–210. PubMed Abstract | Publisher Full Text\n\nWoldt E, Sebti Y, Solt LA, et al.: Rev-erb-α modulates skeletal muscle oxidative capacity by regulating mitochondrial biogenesis and autophagy. Nat Med. 2013; 19(8): 1039–1046. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBass J, Takahashi JS: Circadian integration of metabolism and energetics. Science. 2010; 330(6009): 1349–1354. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoodman CA, Frey JW, Mabrey DM, et al.: The role of skeletal muscle mTOR in the regulation of mechanical load-induced growth. J Physiol. 2011; 589(Pt 22): 5485–5501. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoike N, Yoo SH, Huang HC, et al.: Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science. 2012; 338(6105): 349–354. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSolt LA, Wang Y, Banerjee S, et al.: Regulation of circadian behaviour and metabolism by synthetic REV-ERB agonists. Nature. 2012; 485(7396): 62–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYagita K, Horie K, Koinuma S, et al.: Development of the circadian oscillator during differentiation of mouse embryonic stem cells in vitro. Proc Natl Acad Sci U S A. 2010; 107(8): 3846–3851. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWolff G, Esser KA: Scheduled exercise phase shifts the circadian clock in skeletal muscle. Med Sci Sports Exerc. 2012; 44(9): 1663–1670. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKondratov RV, Kondratova AA, Gorbacheva VY, et al.: Early aging and age-related pathologies in mice deficient in BMAL1, the core componentof the circadian clock. Genes Dev. 2006; 20(14): 1868–1873. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiller BH, McDearmon EL, Panda S, et al.: Circadian and CLOCK-controlled regulation of the mouse transcriptome and cell proliferation. Proc Natl Acad Sci U S A. 2007; 104(9): 3342–3347. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBunger MK, Wilsbacher LD, Moran SM, et al.: Mop3 is an essential component of the master circadian pacemaker in mammals. Cell. 2000; 103(7): 1009–1017. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcDearmon EL, Patel KN, Ko CH, et al.: Dissecting the functions of the mammalian clock protein BMAL1 by tissue-specific rescue in mice. Science. 2006; 314(5803): 1304–1308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng B, Larkin DW, Albrecht U, et al.: The mPer2 gene encodes a functional component of the mammalian circadian clock. Nature. 1999; 400(6740): 169–173. PubMed Abstract | Publisher Full Text\n\nBae K, Lee K, Seo Y, et al.: Differential effects of two period genes on the physiology and proteomic profiles of mouse anterior tibialis muscles. Mol Cells. 2006; 22(3): 275–284. PubMed Abstract\n\nZhang Y, Fang B, Emmett MJ, et al.: GENE REGULATION. Discrete functions of nuclear receptor Rev-erbα couple metabolism to the clock. Science. 2015; 348(6242): 1488–1492. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo B, Chatterjee S, Li L, et al.: The clock gene, brain and muscle Arnt-like 1, regulates adipogenesis via Wnt signaling pathway. FASEB J. 2012; 26(8): 3453–3463. PubMed Abstract | Publisher Full Text\n\nPanda S, Antoch MP, Miller BH, et al.: Coordinated transcription of key pathways in the mouse by the circadian clock. Cell. 2002; 109(3): 307–320. PubMed Abstract | Publisher Full Text\n\nRudic RD, McNamara P, Curtis AM, et al.: BMAL1 and CLOCK, two essential components of the circadian clock, are involved in glucose homeostasis. PLoS Biol. 2004; 2(11): e377. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchiaffino S, Dyar KA, Ciciliot S, et al.: Mechanisms regulating skeletal muscle growth and atrophy. FEBS J. 2013; 280(17): 4294–4314. PubMed Abstract | Publisher Full Text\n\nZhang X, Patel SP, McCarthy JJ, et al.: A non-canonical E-box within the MyoD core enhancer is necessary for circadian expression in skeletal muscle. Nucleic Acids Res. 2012; 40(8): 3419–3430. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShavlakadze T, Anwari T, Soffe Z, et al.: Impact of fasting on the rhythmic expression of myogenic and metabolic factors in skeletal muscle of adult mice. Am J Physiol Cell Physiol. 2013; 305(1): C26–35. PubMed Abstract | Publisher Full Text\n\nRudnicki MA, Schnegelsberg PN, Stead RH, et al.: MyoD or Myf-5 is required for the formation of skeletal muscle. Cell. 1993; 75(7): 1351–1359. PubMed Abstract | Publisher Full Text\n\nBuckingham M: Myogenic progenitor cells and skeletal myogenesis in vertebrates. Curr Opin Genet Dev. 2006; 16(5): 525–532. PubMed Abstract | Publisher Full Text\n\nJanich P, Pascual G, Merlos-Suárez A, et al.: The circadian molecular clock creates epidermal stem cell heterogeneity. Nature. 2011; 480(7376): 209–214. PubMed Abstract | Publisher Full Text\n\nKarpowicz P, Zhang Y, Hogenesch JB, et al.: The circadian clock gates the intestinal stem cell regenerative state. Cell Rep. 2013; 3(4): 996–1004. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPlikus MV, Vollmers C, de la Cruz D, et al.: Local circadian clock gates cell cycle progression of transient amplifying cells during regenerative hair cycling. Proc Natl Acad Sci U S A. 2013; 110(23): E2106–2115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCossu G, Borello U: Wnt signaling and the activation of myogenesis in mammals. EMBO J. 1999; 18(24): 6867–6872. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBentzinger CF, Wang YX, von Maltzahn J, et al.: Fibronectin regulates Wnt7a signaling and satellite cell expansion. Cell stem cell. 2013; 12(1): 75–87. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLe Grand F, Jones AE, Seale V, et al.: Wnt7a activates the planar cell polarity pathway to drive the symmetric expansion of satellite stem cells. Cell stem cell. 2009; 4(6): 535–547. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCarthy JJ, Mula J, Miyazaki M, et al.: Effective fiber hypertrophy in satellite cell-depleted skeletal muscle. Development. 2011; 138(17): 3657–3666. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLipton JO, Yuan ED, Boyle LM, et al.: The Circadian Protein BMAL1 Regulates Translation in Response to S6K1-Mediated Phosphorylation. Cell. 2015; 161(5): 1138–1151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeeney KA, Hansen LL, Putker M, et al.: Daily magnesium fluxes regulate cellular timekeeping and energy balance. Nature. 2016; 532(7599): 375–379. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLatres E, Amini AR, Amini AA, et al.: Insulin-like growth factor-1 (IGF-1) inversely regulates atrophy-induced genes via the phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin (PI3K/Akt/mTOR) pathway. J Biol Chem. 2005; 280(4): 2737–2744. PubMed Abstract | Publisher Full Text\n\nBentzinger CF, Lin S, Romanino K, et al.: Differential response of skeletal muscles to mTORC1 signaling during atrophy and hypertrophy. Skelet Muscle. 2013; 3(1): 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurke L, Downes M, Carozzi A, et al.: Transcriptional repression by the orphan steroid receptor RVR/Rev-erb beta is dependent on the signature motif and helix 5 in the E region: functional evidence for a biological role of RVR in myogenesis. Nucleic Acids Res. 1996; 24(18): 3481–3489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRaichur S, Fitzsimmons RL, Myers SA, et al.: Identification and validation of the pathways and functions regulated by the orphan nuclear receptor, ROR alpha1, in skeletal muscle. Nucleic Acids Res. 2010; 38(13): 4296–4312. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLau P, Bailey P, Dowhan DH, et al.: Exogenous expression of a dominant negative RORalpha1 vector in muscle cells impairs differentiation: RORalpha1 directly interacts with p300 and myoD. Nucleic Acids Res. 1999; 27(2): 411–420. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPircher P, Chomez P, Yu F, et al.: Aberrant expression of myosin isoforms in skeletal muscles from mice lacking the rev-erbAalpha orphan receptor gene. Am J Physiol Regul Integr Comp Physiol. 2005; 288(2): R482–490. PubMed Abstract | Publisher Full Text\n\nAzmi S, Ozog A, Taneja R: Sharp-1/DEC2 inhibits skeletal muscle differentiation through repression of myogenic transcription factors. J Biol Chem. 2004; 279(50): 52643–52652. PubMed Abstract | Publisher Full Text\n\nLing BM, Gopinadhan S, Kok WK, et al.: G9a mediates Sharp-1-dependent inhibition of skeletal muscle differentiation. Mol Biol Cell. 2012; 23(24): 4778–4785. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoenneberg T, Allebrandt KV, Merrow M, et al.: Social jetlag and obesity. Curr Biol. 2012; 22(10): 939–943. PubMed Abstract | Publisher Full Text\n\nDeschenes MR, Sharma JV, Brittingham KT, et al.: Chronobiological effects on exercise performance and selected physiological responses. Eur J Appl Physiol Occup Physiol. 1998; 77(3): 249–256. PubMed Abstract | Publisher Full Text\n\nGauthier A, Davenne D, Martin A, et al.: Time of day effects on isometric and isokinetic torque developed during elbow flexion in humans. Eur J Appl Physiol. 2001; 84(3): 249–252. PubMed Abstract | Publisher Full Text\n\nNicolas A, Gauthier A, Bessot N, et al.: Time-of-day effects on myoelectric and mechanical properties of muscle during maximal and prolonged isokinetic exercise. Chronobiol Int. 2005; 22(6): 997–1011. PubMed Abstract | Publisher Full Text\n\nPearson SJ, Onambele GN: Influence of time of day on tendon compliance and estimations of voluntary activation levels. Muscle Nerve. 2006; 33(6): 792–800. PubMed Abstract | Publisher Full Text\n\nSedliak M, Finni T, Cheng S, et al.: Diurnal variation in maximal and submaximal strength, power and neural activation of leg extensors in men: multiple sampling across two consecutive days. Int J Sports Med. 2008; 29(3): 217–224. PubMed Abstract | Publisher Full Text\n\nReilly T, Waterhouse J: Sports performance: is there evidence that the body clock plays a role? Eur J Appl Physiol. 2009; 106(3): 321–332. PubMed Abstract | Publisher Full Text\n\nDrust B, Waterhouse J, Atkinson G, et al.: Circadian rhythms in sports performance--an update. Chronobiol Int. 2005; 22(1): 21–44. PubMed Abstract | Publisher Full Text\n\nSedliak M, Finni T, Cheng S, et al.: Effect of time-of-day-specific strength training on serum hormone concentrations and isometric strength in men. Chronobiol Int. 2007; 24(6): 1159–1177. PubMed Abstract | Publisher Full Text\n\nSouissi N, Sesboüé B, Gauthier A, et al.: Effects of one night's sleep deprivation on anaerobic performance the following day. Eur J Appl Physiol. 2003; 89(3–4): 359–366. PubMed Abstract | Publisher Full Text\n\nBambaeichi E, Reilly T, Cable NT, et al.: Influence of time of day and partial sleep loss on muscle strength in eumenorrheic females. Ergonomics. 2005; 48(11–14): 1499–1511. PubMed Abstract | Publisher Full Text\n\nZambon AC, McDearmon EL, Salomonis N, et al.: Time- and exercise-dependent gene regulation in human skeletal muscle. Genome Biol. 2003; 4(10): R61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMurphy BA, Wagner AL, McGlynn OF, et al.: Exercise influences circadian gene expression in equine skeletal muscle. Vet J. 2014; 201(1): 39–45. PubMed Abstract | Publisher Full Text\n\nNakao R, Yamamoto S, Horikawa K, et al.: Atypical expression of circadian clock genes in denervated mouse skeletal muscle. Chronobiol Int. 2015; 32(4): 486–496. PubMed Abstract | Publisher Full Text\n\nCastillo C, Molyneux P, Carlson R, et al.: Restricted wheel access following a light cycle inversion slows re-entrainment without internal desynchrony as measured in Per2Luc mice. Neuroscience. 2011; 182: 169–176. PubMed Abstract | Publisher Full Text\n\nKaratsoreos IN, Bhagat S, Bloss EB, et al.: Disruption of circadian clocks has ramifications for metabolism, brain, and behavior. Proc Natl Acad Sci U S A. 2011; 108(4): 1657–1662. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPan A, Schernhammer ES, Sun Q, et al.: Rotating night shift work and risk of type 2 diabetes: two prospective cohort studies in women. PLoS Med. 2011; 8(12): e1001141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParkes KR: Shift work and age as interactive predictors of body mass index among offshore workers. Scand J Work Environ Health. 2002; 28(1): 64–71. PubMed Abstract | Publisher Full Text\n\nScheer FA, Hilton MF, Mantzoros CS, et al.: Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci U S A. 2009; 106(11): 4453–4458. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTurek FW, Joshu C, Kohsaka A, et al.: Obesity and metabolic syndrome in circadian Clock mutant mice. Science. 2005; 308(5724): 1043–1045. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeFronzo RA, Tripathy D: Skeletal muscle insulin resistance is the primary defect in type 2 diabetes. Diabetes Care. 2009; 32(Suppl 2): S157–163. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDyar KA, Ciciliot S, Wright LE, et al.: Muscle insulin sensitivity and glucose metabolism are controlled by the intrinsic muscle clock. Mol Metab. 2013; 3(1): 29–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHodge BA, Wen Y, Riley LA, et al.: The endogenous molecular clock orchestrates the temporal separation of substrate metabolism in skeletal muscle. Skelet Muscle. 2015; 5: 17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuez H, Staels B: The nuclear receptors Rev-erbs and RORs integrate circadian rhythms and metabolism. Diab Vasc Dis Res. 2008; 5(2): 82–88. PubMed Abstract | Publisher Full Text\n\nYin L, Wu N, Lazar MA: Nuclear receptor Rev-erbalpha: a heme receptor that coordinates circadian rhythm and metabolism. Nucl Recept Signal. 2010; 8: e001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDelezie J, Dumont S, Dardente H, et al.: The nuclear receptor REV-ERBα is required for the daily balance of carbohydrate and lipid metabolism. FASEB J. 2012; 26(8): 3321–3335. PubMed Abstract | Publisher Full Text\n\nCho H, Zhao X, Hatori M, et al.: Regulation of circadian behaviour and metabolism by REV-ERB-α and REV-ERB-β. Nature. 2012; 485(7396): 123–127. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLau P, Nixon SJ, Parton RG, et al.: RORalpha regulates the expression of genes involved in lipid homeostasis in skeletal muscle cells: caveolin-3 and CPT-1 are direct targets of ROR. J Biol Chem. 2004; 279(35): 36828–36840. PubMed Abstract | Publisher Full Text\n\nSteinmayr M, André E, Conquet F, et al.: staggerer phenotype in retinoid-related orphan receptor alpha-deficient mice. Proc Natl Acad Sci U S A. 1998; 95(7): 3960–3965. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLamia KA, Sachdeva UM, DiTacchio L, et al.: AMPK regulates the circadian clock by cryptochrome phosphorylation and degradation. Science. 2009; 326(5951): 437–440. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShi S, Hida A, McGuinness OP, et al.: Circadian clock gene Bmal1 is not essential; functional replacement with its paralog, Bmal2. Curr Biol. 2010; 20(4): 316–321. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKennaway DJ, Owens JA, Voultsios A, et al.: Metabolic homeostasis in mice with disrupted Clock gene expression in peripheral tissues. Am J Physiol Regul Integr Comp Physiol. 2007; 293(4): R1528–1537. PubMed Abstract | Publisher Full Text\n\nBugge A, Feng D, Everett LJ, et al.: Rev-erbα and Rev-erbβ coordinately protect the circadian clock and normal metabolic function. Genes Dev. 2012; 26(7): 657–667. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGachon F, Olela FF, Schaad O, et al.: The circadian PAR-domain basic leucine zipper transcription factors DBP, TEF, and HLF modulate basal and inducible xenobiotic detoxification. Cell Metab. 2006; 4(1): 25–36. PubMed Abstract | Publisher Full Text\n\nFranken P, Lopez-Molina L, Marcacci L, et al.: The transcription factor DBP affects circadian sleep consolidation and rhythmic EEG activity. J Neurosci. 2000; 20(2): 617–625. PubMed Abstract"
}
|
[
{
"id": "14726",
"date": "14 Jul 2016",
"name": "R. Daniel Rudic",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract Given the clock as IS an evolutionarily-conserved time-keeping mechanism that synchronizes internal physiology to environmental cues, locomotor activities initiated by skeletal muscle enable entrainment to the light-dark cycles on earth, thus ensuring organismal survival and fitness.\n\nThe statement locomotor activities initiated by skeletal muscle is a tenuous statement. Locomotor activities are most likely secondary to central clock function, as was shown in the McDearmon article which showed that muscle specific rescue did not restore locomotor rhythm.\n\nIntroduction As locomotor activity, the essential function of skeletal muscle in all animal species is under direct circadian clock control through sleep-wake cycles, and the intimate interplay between clock and skeletal muscle physiology is evolutionarily-conserved to ensure fitness and survival.\nRun on sentence\nThe tissue-intrinsic circadian clock in skeletal muscle This hierarchal machinery is composed of a central pacemaker in the brain’s SCN and peripheral clocks in nearly every tissue and cell types, driven by the central clock pacemaker under normal physiological conditions. (repetitive as mentioned in first paragraph)\n\n……the brain’s SCN and peripheral clocks in nearly every tissue and cell types\ncell type repetition of ‘intricate’\n\nAccumulating evidence indicates an intimate interplay between circadian clock machinery and metabolic regulations, either at the level of temporal control evident in many key metabolic processes in distinct metabolic tissues, or in the maintenance of whole-body metabolic homeostasis Overuse of ‘metabolic’ and metabolic regulations should be regulation\nClock modulation of muscle growth, repair and mass maintenance as early as in 8-weeks old mice\n8 week\n\nprolong survival of Bmal1-null animals, whereas brain-specific rescue was not sufficient\nbrain rescue improved survival to 75% in the length of the experiment, saying not sufficient overstated.\nDespite discussion of implications to skeletal muscle disorders, specific links to which skeletal muscle diseases may be implicated are lacking.",
"responses": []
},
{
"id": "15272",
"date": "28 Jul 2016",
"name": "Lei Yin",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe review article summarizes the current understanding about the molecular circadian network for its role in muscle development and muscle biology. This topic is relatively new but will have significant implications in exercise, metabolism and aging. The manuscript is well-written and easy to read. I fully support its indexation in its current format.\nI have a few minor suggestions to help further strengthen the manuscript:\nIt is expected that myocyte-specific molecular clock controls the diurnal expression of key genes that are important for muscle function, such as MyoD. It is less known whether molecular circadian clock within myocytes could directly control the key signalling pathways of muscle metabolism. It would be great if the authors could elaborate on this topic.\n\nThe question regarding the regulation of muscle circadian clock is missing in the current manuscript. Do we know anything about hormonal or nutritional dependent regulation of muscle circadian clock? Does muscle circadian clock change during obesity, diabetes and the aging process? These knowledge will enhance the readability of this manuscript.",
"responses": []
},
{
"id": "15134",
"date": "01 Aug 2016",
"name": "Henrik Oster",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting review paper summarising the current knowledge about the role of the circadian clock in skeletal muscle development and function.\nI have few comments which might help to improve the paper:\nIntro: last two sentences are difficult to understand. \"As locomotor activity...\" this is not a complete sentence / \"It is therefore possible...\" - consider shortening. Is your main point to say that there is still a lot to do?\n\nChapter \"The tissue-intrinsic circadian clock in skeletal muscle\"\n- Daily oscillations are not GOVERNED by outside rhythms; consider ENTRAINED\n- Peripheral clocks are not DRIVEN by the SCN; consider COORDINATED/RESET.\n\nCarefully check spelling of genes and proteins, in particular capitalisations\n\nConsider discussing the role of metabolic feedback on SM clock regulation and function.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1549
|
https://f1000research.com/articles/5-1548/v1
|
30 Jun 16
|
{
"type": "Opinion Article",
"title": "Revitalizing biomedical research: recommendations from the Future of Research Chicago Symposium",
"authors": [
"Kyle T. Dolan",
"Joseph F. Pierre",
"Erin J. Heckler",
"Joseph F. Pierre",
"Erin J. Heckler"
],
"abstract": "The biomedical research enterprise faces considerable structural challenges after years of stagnant funding coupled with steady growth of the pool of graduate students and postdoctoral scientists. Input from junior scientists into the nature of how these challenges affect both the quality of the enterprise and career outcomes is essential to craft effective reforms that will bring a new era of robustness into biomedical research. In October 2015, junior scientists based in Chicago organized the Future of Research Chicago Symposium. The goals of the meeting were twofold: first, to educate the local community about structural problems in biomedical science; and second, to survey scientists in the Midwest, particularly postdocs, in order to find out their views on these issues and solicit suggestions for improvement. We present the recommendations of Symposium participants as distilled by the organizers. These recommendations reflect junior scientists’ desire for diversification of career development opportunities within the framework of doctoral and postdoctoral training and for policies at funding agencies that demonstrate a stronger commitment to supporting trainees and new investigators. We discuss practical steps that can be taken to enable these reforms, highlighting the responsibilities of junior scientists, faculty, funding agencies, and other stakeholders in working toward the goal of a revitalized biomedical research system.",
"keywords": [
"Biomedical research",
"junior scientists",
"training",
"funding",
"mentoring",
"career development",
"culture of science"
],
"content": "Introduction\n\nThe biomedical research enterprise in the United States faces significant structural challenges1. Since 2000, federal funding for biomedical research and development as a percentage of GDP has declined as Congressional appropriations continually lag behind inflation2. Over the past 20 years, the percentage of grant applications funded by the National Institutes of Health (NIH) has fallen by 50%. As funds disappear, biomedical research has become hypercompetitive, with the unfortunate result of fewer biomedical PhDs working in their field of training and increasing percentages leaving research altogether, compared with 25 years ago (Appendix E in 3).\n\nFor those early in their scientific careers, including graduate students, postdoctoral researchers (postdocs), and junior faculty, the pressure of structural problems in the biomedical workforce is especially heavy. Postdoctoral training has come under renewed and particularly intense scrutiny in recent years. Postdocs are an important part of biomedical research, providing most of the skilled labor responsible for scientific discovery and innovation4. Historically, the postdoc position was designed to be a temporary period of focused training needed for future independent research careers. However, since 1990, the estimated postdoc population has doubled, causing an increase in the average length of postdoc training to 5–6 years and resulting in fewer percentages of postdocs obtaining faculty positions. Many commentators have noted that a postdoc position today serves less as a temporary period of advanced training and more as a source of cheap scientific labor5–7, equating to ~$16/hour with few fringe benefits. In addition, as federal funding for research has shrunk, the average age of NIH investigators and the average age at which an investigator first obtains NIH R01 funding have both steadily increased over the past 25 years3. Strikingly, fewer than 3% of NIH funded investigators are now under the age of 363. Together, greater numbers of postdocs without permanent jobs in their academic fields has supported the concept of the ‘postdoc holding tank’.\n\nBeyond economic forces, a lack of preparation or motivation for seeking careers outside academia appears to be contributing to the swelling number of postdocs. The skills acquired during traditional doctoral and postdoctoral training concentrate on preparation for academic careers. However, many career paths for doctorate-level scientists require a broad skill set including competencies in management, business, communications, and leadership that are not a significant focus of the traditional PhD training model3,8. Among PhD students, interest in an academic research career declines during graduate training9,10. However, junior scientists commonly report that faculty advisors often encourage academic research as the definition of “successful” career paths while explicitly discouraging graduate students and postdocs from pursuing non-academic careers9,11.\n\nIn response to employment trends among young biomedical doctorates, the NIH recently initiated a program called Broadening Experiences in Scientific Training (BEST)12 to expand career preparation for biomedical graduate students and postdocs. Between 2013 and 2014, 17 universities across the country received BEST grants, which provide five years of non-renewable funding to develop and test new approaches to career development that would complement traditional graduate and postdoctoral training. While the BEST program has largely been perceived as a success, it is unclear whether these programs will continue once the funding runs out, or whether the approaches developed within the BEST program can be broadly replicated at non-BEST institutions. Without long-term strategies to address the realities of career trends among PhDs, and absent a major shift in the research funding landscape, the outlook for continuing to attract talented young minds to biomedical research remains dim.\n\nIn response to the endemic issues in postdoctoral training, and to the larger problems of the biomedical community as a whole, postdocs and young faculty alike are increasingly engaging in advocacy efforts aimed at changing policies and practices to improve the research environment in the US. A recent notable example is the Future of Research (FoR) Symposium organized by junior scientists in Boston in 2014. A white paper of recommendations produced by FoR11 was recently included in a survey of panel reports aimed at distilling necessary reforms for biomedical research13.\n\nInspired by the events in Boston and motivated by FoR’s stated desire to see junior scientists advocate for change spread throughout the country11, a group of postdocs in Chicago organized their own meeting, the Future of Research Chicago Symposium (FoR Chicago), which took place on October 29, 2015 (for meeting information, see http://futureofresearch.org/chicago/). In this report, we, the organizers of FoR Chicago, lay out several major issues for reform identified by our Symposium. We summarize the solutions generated by open discussions among meeting participants and offer specific recommendations for achieving these reforms.\n\n\nPart I: Guest speakers at FoR Chicago\n\nThe FoR Chicago Symposium opened with keynote lectures designed to introduce policy issues in the biomedical research ecosystem as context for participant-centered discussions later in the day. The speakers were chosen for their expertise in the policy of science, as well as to represent a diverse group of stakeholders. Nancy Schwartz, Dean for Postdoctoral Affairs at the University of Chicago, opened the day with remarks framing the topics of biomedical research sustainability and how to achieve better outcomes for scientists in training. Keith Yamamoto (UCSF), who served on the NIH task force charged with reviewing the biomedical research workforce3, delivered the first keynote lecture. According to Dr. Yamamoto, the PhD-to-postdoc transition is currently viewed as a relatively fixed part of scientific training; the postdoctoral period constitutes a “hub” from which trainees may then follow numerous career paths. Dr. Yamamoto proposed that graduate school should be the “hub” from which a newly minted PhD could opt to pursue many career paths. This would be more sensible, said Dr. Yamamoto, because many careers do not require postdoctoral training; furthermore, such a system would help control the growth of the postdoc population. Following Dr. Yamamoto, Greg Petsko (Cornell), who chaired the National Academies’ Committee to Review the State of Postdoctoral Experience in Scientists and Engineers4, gave the second keynote talk. Dr. Petsko recounted the study, which led to the National Academies’ 2014 report, “The Postdoctoral Experience Revisited”. After presenting highlights of the report, Dr. Petsko offered some of his ideas about how the postdoc dilemma fit into larger challenges in the research enterprise related to funding, publishing, and judging high-quality science.\n\nThe keynote lectures were followed by a panel discussion exploring the evolution of scientific training. The panelists were Kay Lund (NIH), Gary McDowell (formerly of Tufts University, now with Future of Research), Mary O’Riordan (University of Michigan), and Krisztina Eleki (Chicago Council on Science and Technology), who discussed the mismatch between the aims of traditional scientific training and the contemporary career training needs of graduate students and postdocs, along with ideas for how to align training goals to better meet these needs. In both the lectures and panel discussion, audience members had the opportunity to ask questions of the speakers; there was also time for informal engagement during the morning coffee break and lunch, where the speakers sat among the attendees for extended conversation.\n\n\nPart II: Participant-centered discussions\n\nIn the afternoon, we invited all the participants of the Symposium—postdocs, students, faculty, staff, and others—to debate key issues and identify potential reforms through a series of moderated workshops. The workshops were organized around the following five topics: Revolutionizing Training, Curricular Reforms/Experiential Learning, Incentivizing Good Science, Funding Mechanisms, and Scientific Workforce Structure. The topics were chosen by the organizers based on key issues identified at the 2014 FoR Boston Symposium as well as similar events taking place in the spring of 2015 at the Universities of Michigan (FOBGAPT; http://www.rackham.umich.edu/fobgapt) and Wisconsin-Madison14. Participants could attend two workshops of their choosing. During the workshops, problems and solutions were written on sticky notes and grouped into relevant categories. Following the event, we compiled suggestions into briefs from each workshop, and then compared all the briefs to identify major themes. There was a notable degree of overlapping ideas among the different workshops, which underscores the interrelatedness of challenges in biomedical research.\n\nIn almost every workshop session, participants expressed concern at a lack of education and training for careers paths apart from academic faculty. This shortfall of career preparation for new PhDs is cited by junior and senior scientists alike as one of the major problems facing biomedical research1,3,10,11,13,15. Because of the pressing nature of this issue, along with data showing that most biomedical PhDs move into careers outside academia, the NIH recently instituted the BEST program to expand career development resources at a selected group of universities who applied12.\n\nAmong participants’ suggested solutions to the problem of narrow career training, we identified a set that broadly calls for the establishment of “professional PhD” programs alongside or in place of the traditional PhD education. PhD programs with a stronger professional development component added as curricular or extracurricular activities would provide earlier exposure to non-academic career options, along with enhanced knowledge and skills needed for these careers. Participants put forth a number of implementation methods to create professional PhD programs including: 1) allowing students to take courses in a professional program of study (e.g. medicine, business, law) as electives, or a part of a “minor concentration” during the PhD; 2) expanding the number of joint degree programs whereby students could simultaneously work toward a PhD and another degree such as an MBA or JD (not counting MD-PhD programs); and 3) adding internship or externship work experience as a regular part of PhD training.\n\nApart from expanding the range of formal academic approaches to diversifying career preparation, participants also expressed a desire for greater development of “soft” skills that are portable and relevant across many careers. To do this, institutions housing trainees could adopt standards for achieving proficiency in various skill sets and require departments and/or PhD programs to provide training in these areas. One such set of standards that already exists for this very purpose is the National Postdoctoral Association Core Competencies (http://www.nationalpostdoc.org/?CoreCompetencies). Furthermore, it was recommended by participants that institutions should solicit input from various career sectors in order to align training goals with the needs of employers.\n\nParticipants also said that students and postdocs should take on more responsibility for career planning and work to achieve a higher level of self-confidence in considering future career opportunities. Because many trainees feel pressure to succeed on the project in front of them and to pursue an academic career, they may view time spent on career planning as a waste or an admission of failure. Many participants cited individual development plans (IDPs) as a useful tool for reflecting on one’s career aspirations and for building confidence in those aspirations. Trainees should involve Principal Investigators (PIs) in creating IDPs to establish a clear set of goals as well as a shared sense of responsibility and accomplishment. There were also a number of people who believed that students and postdocs interested in careers outside academia should be empowered to seek out mentors in those areas to provide guidance complementary to that of their PI.\n\nWhile mentorship was not a specific workshop topic, numerous discussions touched on mentoring as an area in need of significant improvement. Many participants expressed frustration with either a lack of attention or guidance from PIs, or mentoring practices that did not take trainees’ goals into account. Participants felt that improving mentorship would enhance trainee productivity and increase the quality of trainees’ research, as well as promote better development for a range of careers.\n\nSeveral suggestions were put forth to address this issue. First, the availability of mentor coaching programs to build mentorship skills should be increased. We also think that institutions could provide such training, with guidelines established by funding agencies or scientific societies. An example of such a program is the NIH-funded National Research Mentoring Network (NRMN, https://nrmnet.net), which aims to create and disseminate resources for mentoring and professional development that can be used at all stages of a scientist’s career progression. To make PIs more accountable for mentoring, reviews of PI mentoring activities should be included on performance evaluations for both PIs and trainees, and funding agencies should incorporate scoring criteria for mentorship on grants that call for trainee labor.\n\nMost participants in FoR Chicago expressed concern over the lack of data pertaining to the career outcomes of biomedical trainees, a problem which has been noted elsewhere4,11,15–17. Such career data would give trainees a clearer vision of the diversity of careers available to PhDs and could help guide efforts to design new training models. Most participants agreed that PIs, graduate programs, and/or departments should gather career outcome data on their alumni and make this information publicly available on the Internet.\n\nCollecting employment data will require cooperation between training institutions and the places that hire biomedical PhDs. At the moment, there is little national career data on PhDs; however, participants suggested that such data could be recovered from LinkedIn. A recent report by Silva et al.17 demonstrates the feasibility of this approach for tracking postdocs from a single institution, although it is unclear how this might scale to collect nationwide employment data. Another suggested approach would involve incentivizing companies to track and share PhD hiring trends. Labor data clearinghouses such as Economic Modeling Systems, Inc. (EMSI; economicmodeling.com) have provided valuable information to scientists and labor economists studying the landscape of post-PhD employment8.\n\nThe number of postdocs in US biomedical science is unknown; estimates range from almost 40,000 to over 100,0004,18. In agreement with other reports4,8,11, participants said there are too many postdocs and believed that steps should be taken to reduce their numbers. Many participants agreed that replacing postdocs with permanent, higher-paying staff scientist positions would curtail the growth of the postdoc population while providing a source of skilled scientific labor. This recommendation has been echoed by many commentators13. At the same time, participants desired that the purpose and expectations for postdoctoral training be more clearly defined. Doing so would help eliminate the widespread notion that the postdoc is a “default” step in the career progression of PhD scientists4,19.\n\nAmong our participants, there was a sense that recent declines in funding for science, especially as a part of federal spending, had suppressed the research community’s ability to do important, innovative work. According to participants, scientists should advocate for sustainable federal research budgets. Similar recommendations have been made by many groups1,11,13. Additionally, participants favored cost-saving measures that would create greater returns on funding investment. These included directing more funds toward shared equipment or establishing core facilities that could serve multiple laboratories. In order to encourage long-term planning and reduce wasteful spending, participants recommended that the federal government and NIH should lift restrictions on carrying over unused money between funding years. Furthermore, funding agencies should conduct audits of grant spending upon application for renewal.\n\nMany participants felt concerned that the current process of grant review is too political, with established PIs holding a significant advantage over junior scientists to win funding. They also expressed concerns on grant decisions being biased by scientific pedigree, institutional affiliation, and a heavy reliance on journal metrics as a substitute for quality of ideas. Participants put forward a number of suggestions to improve grant review and help support early-career scientists. The NIH could designate more money for the existing postdoctoral fellowship and career development awards, like the F32 and K award programs. F32 awards support training for individual postdocs with the potential to become independent investigators, while K awards are career development grants for advanced postdocs to move into a PI role, either under the guidance of a mentor (K01) or with more independence (K99). Expanding these mechanisms, particularly the K award program, could provide crucial financial support for new generations of young scientists, and help stabilize the growing average age of NIH funded investigators. However, it was not specified by our participants whether the additional money should support more of these awards or be used to increase their value. Grants originating from early-career scientists should be reviewed separately from mid-career and senior PIs; furthermore, applicant names and institutions should be removed from grant applications to reduce reviewer bias. Creating new block grant mechanisms to fund institutions instead of individual investigators was also put forth as a way to ensure a more equitable funding landscape.\n\nParticipants suggested that new PIs should be cultivated for their fresh, innovative ideas. To support this process, study sections should include junior faculty, and funding agencies could expand award programs that support pilot studies or high-risk, high-reward projects that can provide the basis for a prolonged and productive scientific career.\n\nThe process of grant writing was viewed widely as cumbersome and burdened by regulations. Participants favored several reforms in this area, including standardizing grant application formats and scoring metrics across federal funding agencies; automating and streamlining the application process; and creating or expanding positions for grant support staff working in research institutions that receive federal support. These changes could relieve some of the time pressure demands on PIs who are writing grants, allowing them to focus more on doing science and mentoring trainees.\n\nOur participants, 79% of whom self-identified as trainees, wanted to change the culture of academia to foster robust discovery and innovation and to strengthen the feeling of being part of a scientific community. As a way for scientists to promote innovation (apart from the funding tactics described above), one intriguing suggestion was that major scientific meetings devote a session to highlighting innovative early-career work. Participants called for several actions concerning publishing and data sharing to enable new discoveries and empower the free exchange of ideas among scientists. These included expanding open-access journals, creating open data repositories, and embracing publication of negative results as part of the scientific process.\n\nFinally, participants said that finding joy in one’s work and having self-confidence were critical ingredients for succeeding in science. Unfortunately, they viewed the current culture of science as depriving too many young people of experiencing powerful positive feelings about their work. As a solution, participants recommended that institutions should dedicate greater resources to promoting social cohesion among young scientists, noting that connection with one’s peers can help people remember the joys of science.\n\n\nConclusions\n\nOur vision for the Future of Research Symposium in Chicago was to give a diverse group of scientists, 79% of whom were PhD students and postdocs, a chance to speak out on where reforms are needed in the American biomedical research enterprise and to hear from national leaders. The proposals put forth by the Symposium participants point to several areas requiring critical attention. Scientific training for PhDs and postdocs must be recalibrated to prepare scientists to follow multiple career paths beyond the traditional route in the academy. The postdoctoral training period in particular needs to be revitalized. Young scientists should not have to treat a postdoc position as a default step on their career path, nor should PIs treat it as a source of inexpensive scientific labor. Funding agencies including the NIH and others could take steps to lend greater support to young investigators and encourage bold, innovative research initiatives. And, the scientific community should adopt cultural practices that foster a spirit of openness, collaboration, and community that will help sustain the research enterprise and attract future generations of scientists into its ranks.\n\nSeveral of the recommendations put forth here echo those made by other commentators. Our proposals add to numerous calls for diversifying career training for students and postdocs, and increasing the number of permanent staff scientists employed in biomedical research laboratories1,4,11,13–15,20,21. We also highlight the need for improved mentorship in academia, which reflects a common view among postdocs11 and other groups studying the contemporary trainee experience in biomedical research3,4,8,10,17. We hope that the recommendations provided here will encourage further discussion among scientists and stakeholders from many corners of the biomedical research enterprise and the larger scientific community, thereby adding momentum to the movement for broad reform of the current system.\n\nTranslating these ideas into action will require a concerted effort from stakeholders across the research landscape. Recent examples of such teamwork include the ASAPBio meeting (asapbio.org) featuring junior and senior scientists, journal editors, and funding agency representatives; the ASBMB Sustainability Summit (http://policy.asbmb.org/2016/02/01/the-asbmb-sustainability-summit/), organized by that society’s public policy branch; and the work of organizations such as the National Postdoctoral Association (http://www.nationalpostdoc.org), Rescuing Biomedical Research (http://rescuingbiomedicalresearch.org), and Future of Research (http://futureofresearch.org). We hope that these efforts continue to involve people from all levels of science, including junior scientists. In order to achieve reforms, groups such as these must not overlook the need to win support from rank-and-file faculty across the US. Indeed, we believe that faculty will play a critical role, perhaps the central role, in determining whether the efforts to reform scientific research succeed or fail. In our research system, faculties embody the culture of science; they train young scientists; select which grants to fund; determine which papers should be published, and where; and create institutional policy. Changing the culture of research will require them to buy into the vision of those who would reform it. This is an enormous task, but one that cannot be avoided if we are to create a better future for science and for scientists.",
"appendix": "Author contributions\n\n\n\nK.T.D. drafted the manuscript. All authors were involved in revising the initial draft of the manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nK.T.D. is a member of the board of directors of Future of Research. The views expressed by the authors do not necessarily reflect those of any institutions or organizations with which they are affiliated.\n\n\nGrant information\n\nFunding for the Future of Research Chicago Symposium was provided by the following sponsors: eLife; Garnett-Powers & Associates; GE Healthcare; The National Academy of Sciences; New England BioLabs; The Feinberg School of Medicine, The Graduate School, and Office of Research at Northwestern University; The University of Chicago Medicine & Biological Sciences including the myCHOICE program and NIH Training Grants T32DK007074 and T32GM099697, The University of Chicago Comprehensive Cancer Center, The Joseph P. Kennedy Jr. Intellectual and Developmental Disabilities Research Center, and the Departments of Biochemistry and Molecular Biology, and Pediatrics; The University of Chicago Institute for Molecular Engineering; and the College of Applied Health Sciences, College of Pharmacy, and Honors College of the University of Illinois at Chicago. The myCHOICE program is funded through an NIH BEST grant (DP7OD020316).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe gratefully acknowledge the work of the postdoc volunteers from University of Chicago, Northwestern University, University of Illinois at Chicago, and Rush University, without which the Future of Research Chicago Symposium would not have been possible. We also thank the many faculty and administrators who served as advisors to the organizing team. A full listing of event volunteers and advisors may be found at futureofresearch.org/Chicago. Special thanks to Dr. Rianne Ellenbroek, co-director of FoR Chicago, for her dedication to this event. We thank Gary McDowell, Connie Lee, and Nancy Schwartz for helpful comments on this manuscript.\n\n\nReferences\n\nAlberts B, Kirschner MW, Tilghman S, et al.: Rescuing US biomedical research from its systemic flaws. Proc Natl Acad Sci U S A. 2014; 111(16): 5773–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNational Institutes of Health: NIH Budget Office Price Indexes. 2016. Reference Source\n\nNational Institutes of Health: Biomedical Research Workforce Working Group Report. Bethesda, MD; 2012. Reference Source\n\nNational Academy of Science, National Academy of Engineering, Institute of Medicine: The Postdoctoral Experience Revisited. National Academies Press, Washington, DC; 2014. PubMed Abstract | Publisher Full Text\n\nStephan P: How to Exploit Postdocs. Bioscience. 2013; 63(4): 245–246. Publisher Full Text\n\nBourne HR: A fair deal for PhD students and postdocs. eLife. 2013; 2: e01139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPolka J: Where Will A Biology PhD Take You? 2014. Reference Source\n\nMason JL, Johnston E, Berndt S, et al.: Labor and skills gap analysis of the biomedical research workforce. FASEB J. 2016; pii: fj.201500067R. PubMed Abstract | Publisher Full Text\n\nSauermann H, Roach M: Science PhD career preferences: levels, changes, and advisor encouragement. PLoS One. 2012; 7(5): e36307. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGibbs KD Jr, McGready J, Griffin K: Career Development among American Biomedical Postdocs. CBE Life Sci Educ. 2015; 14(4): ar44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcDowell GS, Gunsalus KT, MacKellar DC, et al.: Shaping the Future of Research: a perspective from junior scientists [version 2; referees: 2 approved]. F1000 Res. 2015; 3: 291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyers FJ, Mathur A, Fuhrmann CN, et al.: The origin and implementation of the Broadening Experiences in Scientific Training programs: an NIH common fund initiative. FASEB J. 2016; 30(2): 507–14. PubMed Abstract | Publisher Full Text\n\nPickett CL, Corb BW, Matthews CR, et al.: Toward a sustainable biomedical research enterprise: Finding consensus and implementing recommendations. Proc Natl Acad Sci U S A. 2015; 112(35): 10832–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKimble J, Bement WM, Chang Q, et al.: Strategies from UW-Madison for rescuing biomedical research in the US. eLife. 2015; 4: e09305. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDaniels RJ: A generation at risk: Young investigators and the future of the biomedical workforce. Proc Natl Acad Sci U S A. 2015; 112(2): 313–318. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPolka JK, Krukenberg KA, McDowell GS: A call for transparency in tracking student and postdoc career outcomes. Mol Biol Cell. 2015; 26(8): 1413–1415. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSilva EA, Des Jarlais C, Lindstaedt B, et al.: Tracking Career Outcomes for Postdoctoral Scholars: A Call to Action. PLoS Biol. 2016; 14(5): e1002458. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarrison HH, Justement LB, Gerbi SA: Biomedical science postdocs: an end to the era of expansion. FASEB J. 2016; 30(1): 41–44. PubMed Abstract | Publisher Full Text\n\nSauermann H, Roach M: SCIENTIFIC WORKFORCE. Why pursue the postdoc path? Science. 2016; 352(6286): 663–4. PubMed Abstract | Publisher Full Text\n\nAmerican Academy of Arts and Sciences: Restoring the Foundation: The Vital Role of Research in Preserving the American Dream. American Academy of Arts and Sciences, Cambridge, MA; 2014. Reference Source\n\nFederation of American Societies for Experimental Biology: Sustaining Discovery in Biological and Medical Sciences. Bethesda, MD; 2015. Reference Source"
}
|
[
{
"id": "14710",
"date": "14 Jul 2016",
"name": "Christopher L. Pickett",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe meeting report “Revitalizing biomedical research: recommendations from the Future of Research Chicago Symposium” from Dolan, et al., describes a meeting in Chicago in 2015 featuring two keynote speakers, a panel session and multiple breakout sessions attended by trainees, staff and faculty. The majority of this report focuses on the recommendations and outcomes of the breakout sessions.\n\nThis paper is an important contribution to the growing literature on how postdocs view their place in the biomedical research enterprise. That said, the authors can make several significant improvements to this manuscript to clarify the outcomes of this meeting and what actions should be taken next.\n\nPrimary recommendations\nThe manuscript would benefit from a display item cataloguing the many recommendations made pertaining to each of the themes.\n\nThe authors should state clearly how the outcomes of their meeting relate to (1) McDowell, et al., (2014) and (2) other reports recommending changes to how students and postdocs are trained.\nPostdocs play a pivotal role in the research enterprise, but this is only the second publication from a postdoc meeting discussing the ailments of the research enterprise. The growth of the postdoc voice in the conversation is an important point that the authors should stress. The authors should clearly state how their outcomes are the same or different from the McDowell paper and what conclusions they draw from these similarities and differences. Many reports have been written discussing trainee issues, and the authors do a nice job referencing much of the literature. The authors should go further to clearly define how their recommendations contribute to the ongoing discussion and highlight areas of agreement or disagreement with these reports.\n\nIt is important for the authors to make very clear for the reader when they are reporting the participants’ recommendations or their own suggestions. In the final paragraph of Theme 1, the authors use “We think…” when discussing recommendations made by participants. This is quite confusing. Other sentences throughout the manuscript should also be clarified to ensure the reader understands who is making the recommendations. It may be useful to move the authors’ suggestions and recommendations to the conclusions and stick to participant recommendations in the themes.\n\nThe authors should take more time in the Conclusions to give direction to the recommendations made at their meeting. Many of the recommendations had been made by others, as the authors indicated, but what does this mean about the progress of change? Which recommendations do the authors think are the most important to focus on? Where, given developments since the end of their meeting, should efforts be focused to make the most change?\n\nSecondary recommendations\nThe authors list that “postdocs, students, faculty, staff, and others” participated in their event. Rough estimates of the percent attendance by career stage would be important to understand the tenor of the recommendations being made. Furthermore, if the data are available, it is important to communicate if one career group, like postdocs, had more of a say in the recommendations of one theme than other groups.\n\nIt is not clear what the authors are trying to convey with this sentence: “Without long-term strategies to address the realities of career trends among PhDs, and absent a major shift in the research funding landscape, the outlook for continuing to attract talented young minds to biomedical research remains dim.” What are “long-term strategies to address the realities of career trends” and how do they help recruit young scientists into the enterprise? Would a “major shift in the research funding landscape” do more to retain young scientists in the enterprise or recruit them to it? This sentence should be recast to capture the authors true intentions or deleted altogether.\n\nThe participants in Theme 1 did not offer “implementation methods” as the authors suggest. Implementation requires a plan to make the recommended change a reality, such as who should be tasked with advocating for and making the change and a timeline for the change to take effect. This type of plan is not offered in the manuscript. Rather, the participants make recommendations for alternative models to achieve the objective, and they should be characterized as such.\n\nIn the final paragraph of Theme 1, the authors list, “First, the availability of…” but there is no “Second…”. The sentences should be recast to address this issue.\n\nIn Theme 2, to give an alternative, yet complementary, picture of what the Silva et al. group did, the authors should consider referencing The Stanford PhD Alumni Employment Project (http://web.stanford.edu/dept/pres-provost/irds/PhDAlumniEmployment). It could also be useful to reference some of the schools that publish career outcomes information such as UCSF, University of Chicago or Tufts University.\n\nIn Theme 4, the participant’s recommendation suggests a lack of interaction and social cohesion is responsible for people not “experiencing powerful positive feelings about their work.” Is this lack of interaction the root cause of poor morale as identified by the participants? How much of the material in Themes 1-3 play a role in lab morale? The authors should expand and clarify what the participants cited as depressing the lab culture.\n\nThe authors discuss several broad problems facing the research enterprise but sometimes obscure the complexity of the reasons behind them. I recommend recasting these sentences to reflect the complexity of the situation. Specifically:\n“However, since 1990, the estimated postdoc population has doubled, causing an increase in the average length of postdoc training to 5–6 years and…” Linking population growth directly to the length of postdoc training leaves out many important details. Stagnant research funding, the dwindling number of faculty positions and poor training for jobs outside of academia, in addition to population growth, have affected the length of postdoc periods. “Over the past 20 years, the percentage of grant applications funded by the National Institutes of Health (NIH) has fallen by 50%.” While accurate, it is important to note that, in the context of the rest of the paragraph, this is due to a doubling of grant applications rather than a decline in the raw number of grants funded. As funds disappear, biomedical research has become hypercompetitive…” Stagnant funding is a major contributor to the hypercompetitive environment, but so is the growth in the population applying for grant funding and the increase in applications referenced above. The authors state the BEST program was implemented, “In response to employment trends among young biomedical doctorates…” This is partially true, but again, oversimplified. The BEST program was initiated in response to a recommendation in the NIH Biomedical Workforce Working Group report from 2012, continuous advocacy from parts of the research community, the apparent success of a variety of existing university-specific programs, and yes, an analysis of employment trends.",
"responses": []
},
{
"id": "15181",
"date": "25 Jul 2016",
"name": "Richard McGee",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nHaving reviewed the manuscript prior to reading the comments of the other reviewer, I would concur with most of the points raised so I won’t repeat them. In particular, however, the current version makes it difficult for a reader to:\nSeparate new ideas from those in concordance with previous reports, especially those that came out of the Boston meeting of postdocs;\n\nDetermine which conclusions and recommendations were coming from the meeting participants vs. the authors;\n\nDetermine which ideas and recommendations had the strongest consensus behind them vs. novel ideas of one or a few individuals\nSpecific Comments:\nIntroduction\nFirst paragraph – It is not really true that research funds are ‘disappearing’. Stating it this way creates a false sense of rapid decline that is not accurate. The bigger change has been on demand side, not on the supply side, as universities keep applying more pressure to faculty to get grants, decreasing institutional support for faculty salaries, and increasing infrastructure costs. This has led to a big increase in the number of applications for the same research pie.\nSecond paragraph – As written, this gives the impression that the postdoc is solely a route to academic research career whereas it is also a route to other research careers in industry and government labs, although it is clear there are many postdocs who aspire to but who will never achieve a research-based career in any of these sectors.\nLast paragraph – By the end of the introduction, it is still unclear how this event was different than FoR in Boston, whether participants overlapped at all, whether Chicago participants had read the recommendations from Boston, etc.\n\nThe meeting – How many individuals attended and about how many participated in each of the breakout groups? As a reader one can’t determine if the report represents the collective thoughts of a large or small number of postdocs and others.\nPart I – To what degree did these keynotes shape discussion, opinion, and what came out in this report? How were they similar or different from speakers in the Boston event? Essentially, to what degree did the choice and messages lead to or influence the findings and recommendations reported here?\nPart II – Who moderated or led the workshops? Were there any guide questions or structure to them, or was it simply as described – providing opportunities for ideas to be ‘voiced’ via sticky notes which were sorted for themes? Do they have any information on the extent of knowledge of prior publications related to the questions being addressed? This is a potentially important issue if a reader is to determine the degree to which what came out of the meeting was spontaneous ideas which can be triangulated with those previously proposed and published, or ‘commentary’ on those.\nTheme 1 – new training paradigms – Soft Skills – It is not clear what is meant by these as they can be interpreted in many ways. What are the perceived priorities and how might they be taught? Just giving a link to the Core Competencies is not sufficient to know what the participants raised as important or not.\nImproved mentorship is a very vague term and difficult to know what to change. Are there issues that came up other than lack of attention and consideration of trainees’ goals? If so, it would be more valuable to cite them rather than generic better mentorship.\nTheme 3 – Several of the recommendations are actually in place and have been for many years, such as carry-forward of unspent funds from one grant year to another. This is standard practice except in unusual circumstances, only requiring clarification if the amount to be carried forward is more than 25%, and even then it is generally approved with reasonable justification. Also, a great deal of funding is going toward shared resources and core facilities, and if more was to go in that direction it would have to come from some other research line.\nThe recommendation for more F32 and K awards was somewhat hard to grasp given the concerns raised for more emphasis on non-PI careers. Currently, the limiting constraint is the number of tenure-track faculty positions so it is unclear how training more postdocs for these roles will improve the situation. Non-tenure track positions have continued to rise so this is one potential growth area for young scientists, albeit with limited stability.\nThe suggestion that grants from early career scientists should be reviewed separately was actually implemented several years ago for NIH review panels.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1548
|
https://f1000research.com/articles/4-105/v1
|
01 May 15
|
{
"type": "Software Tool Article",
"title": "regionReport: Interactive reports for region-based analyses",
"authors": [
"Leonardo Collado-Torres",
"Andrew E. Jaffe",
"Jeffrey T. Leek",
"Leonardo Collado-Torres",
"Andrew E. Jaffe"
],
"abstract": "regionReport is an R package for generating detailed interactive reports from regions of the genome. The report includes quality-control checks, an overview of the results, an interactive table of the genomic regions and reproducibility information. regionReport can easily be expanded with report templates for other specialized analyses. In particular, regionReport has an extensive report template for exploring derfinder results from annotation-agnostic RNA-seq differential expression analyses.",
"keywords": [
"Report",
"Interactive",
"Reproducibility",
"Genomics",
"Sequencing",
"ChIP-seq",
"RNA-seq",
"Software"
],
"content": "Introduction\n\nMany analyses of genomic data result in regions along the genome that associate with a covariate of interest. These genomic regions can result from identifying differentially bound peaks from ChIP-seq data1, identifying differentially methylated regions (DMRs) from DNA methylation data2, or performing base-resolution differential expression analyses using RNA sequencing data3,4. The genomic regions themselves are commonly stored in a GRanges object from GenomicRanges5 when working with R or the BED file format on the UCSC Genome Browser6, but other information on these regions, for example summary statistics on the magnitude of effects and statistical significance, also provide useful information. The usage of R in genomics is increasingly common due to the usefulness and popularity of the Bioconductor project7, and in the latest version (3.0), 181 unique packages use GenomicRanges for many workflows, demonstrating the widespread utility of identifying and summarizing characteristics of genomic regions.\n\nHere we introduce regionReport which allows users to explore genomic regions of interest through interactive stand-alone HTML reports that can be shared with collaborators. These reports are flexible enough to display plots and quality control checks within a given experiment, but can easily be expanded to include custom visualizations and conclusions. The resulting HTML report emphasizes reproducibility of analyses8 by including all the R code without obstructing the resulting plots and tables. We envision regionReport will provide a useful tool for exploring and sharing genomic region-based results from high throughput genomics experiments.\n\n\nMethods\n\nThe package includes a R Markdown template which is processed using knitr9 and rmarkdown10, then styled using knitrBootstrap11. This package generates a HTML report that includes a series of plots for checking the quality of the results and browsing the table of regions. Each element of the report has a brief explanation, although actual interpretation of the results is dataset- and workflow-dependent. To facilitate navigation a menu is always included, which is useful for users interested in a particular section of the report. Figure 1, panel a shows the menu of the general report for a set of regions with associated p-values. The code for each plot or table is hidden by default and can be shown by clicking on the appropriate toggle as shown in Figure 2.\n\nExample region input, the appropriate regionReport function to use, and menu of the resulting report for the general use case (panel a) and derfinder results (panel b).\n\nView by default (panel a) and after clicking on the “R source” toggle (panel b) for a section of the general report. The full report is available at the supplementary website and includes a toggle to hide/show all the R code.\n\nThis section of the report includes a variety of quality control steps which help the user determine whether the results are sensible. The quality control steps explore:\n\nP-values, Q-values, and FWER adjusted p-values\n\nRegion width\n\nRegion area: sum of single-base level statistics (if available)\n\nMean coverage or other score variables (if available)\n\nA combination of density plots and numerical summaries are used in these quality checks. If there are statistically significant regions, the distributions are compared between all regions and the significant ones. For example, the distribution region widths might have a high density of small values for the global results, but shifted towards higher values for the subset of significant regions.\n\nThe report includes plots to visualize the location of all the regions as well as the significant ones. Differences between them can reveal location biases. The nearest known annotation feature for each region is summarized and visually inspected in the report.\n\nAn interactive table with the top 500 (default) regions is included in this section. This allows the user to sort the region information according to their preferred ranking option. For example, lowest p-value, longest width, chromosome, nearest annotation feature, etc. The table also allows the user to search and subset it interactively. A common use case is when the user wants to check if any of the regions are near a known gene of their interest.\n\nAt the end of the report, detailed information is provided on how the analysis was performed. This includes the actual function call to generate the report, the path where the report was generated, time spent, and the detailed R session information including package versions of all the dependencies.\n\nThe R code for generating the plots and tables in the report is included in the report itself, thus allowing users to manually reproduce any section of the report, customize them, or simply change the graphical parameters to their liking.\n\nWhen exploring derfinder results, for each of the best 100 (default) DERs a plot showing the coverage per sample is included in the report. These plots allow the user to visualize the differences identified by derfinder along known exons, introns and isoforms. The plots are created using derfinderPlot, also available via Bioconductor.\n\nDue to the intrinsic variability in RNA-seq coverage data or mapping artifacts, in situations where there are two candidate DERs that are relatively close there might be reasons to consider them a single candidate DER and its important to visualize them. This tailored report groups candidate DERs into clusters based on a distance cutoff. After ranking them by their area, for the top 20 (default) clusters it plots tracks with the coverage by sample, the mean coverage by group, the identified candidate DERs colored by whether they are statistically significant, and known alternative transcripts. Figure 1, panel b shows the main categories of the report generated from a richer region data set than in the general case.\n\nInstallation. regionReport and required dependencies can be easily installed from Bioconductor with the following commands:\n\nsource(“http://bioconductor.org/biocLite.R”)\n\nbiocLite(“regionReport”)\n\nInput. To generate the report, the user first has to identify the regions of interest according to their analysis workflow. For example, by performing bumphunting to identify DMRs with bumphunter. The report is then created using renderReport() which is the main function in this package as shown in Figure 1, panel a. The argument customCode can be used to customize the report if necessary.\n\nFor the derfinder use case, the derfinderReport() function creates the recommended report that includes visualizations of the coverage information for the best regions and clusters of regions.\n\nOutput. A small example can be generated using:\n\nexample(“renderReport”, “regionReport”, ask=FALSE)\n\nThe resulting HTML file will open in the users default browser. Note that alternative output formats such as PDF files can also be generated, although they are not as dynamic and interactive as the HTML format.\n\n\nUse cases\n\nThe supplementary website contains reports using DiffBind, bumphunter and derfinder results. The derfinder use case is illustrated with data sets previously described in 3 which span simulation results, a moderately sized data set (25 samples), and a large data set with 487 samples; thus covering a wide range of scenarios.\n\n\nSummary\n\nregionReport creates interactive reports from a set of regions and can be used in a wide range of genomic analyses. Reports generated with regionReport can easily be extended to include further quality checks and interpretation of the results specific to the data set under study. These shareable documents are very powerful when exploring different parameter values of an analysis workflow or applying the same method to a wide variety of data sets. The reports allow users to visually check the quality of the results, explore the properties of the genomic regions under study, and inspect the best regions and interactively explore them.\n\nFurthermore, regionReport promotes reproducibility of data exploration and analysis. Each report provides R code that can be used as the starting point for other analyses within a dataset. regionReport provides a flexible output for exploring and sharing results from high throughput genomics experiments.\n\n\nSoftware availability\n\nregionReport is freely available via Bioconductor at bioconductor.org.\n\nhttp://leekgroup.github.io/regionReportSupp/ hosts the code for generating three types of reports as well as the resulting HTML reports generated by regionReport. Versions of all software used are included in the reports.\n\nThe latest source code is available at Bioconductor and github.com/leekgroup/regionReport via the git-svn bridge although we recommend users to install regionReport directly from Bioconductor.\n\nArchived source code available at http://dx.doi.org/10.5281/zenodo.17083\n\nArtistic-2.0.",
"appendix": "Author contributions\n\n\n\nL.C-T. conceived and developed the regionReport package, supervised by A.E.J. and J.T.L. All authors wrote and approved the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nJ.T.L. was partially supported by NIH Grant 1R01GM105705, L.C-T. was supported by Consejo Nacional de Ciencia y Tecnolog’ia México 351535.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nStark R, Brown G: DiffBind: differential binding analysis of ChIP-Seq peak data. 2011. Reference Source\n\nJaffe AE, Murakami P, Lee H, et al.: Bump hunting to identify differentially methylated regions in epigenetic epidemiology studies. Int J Epidemiol. 2012; 41(1): 200–209. Publisher Full Text\n\nTorres LC, Frazee AC, Love MI, et al.: derfinder: Software for annotation-agnostic rna-seq differential expression analysis. bioRxiv. 2015; 015370. Publisher Full Text\n\nFrazee AC, Sabunciyan S, Hansen KD, et al.: Differential expression analysis of RNA-seq data at single-base resolution. Biostatistics. 2014; 15(3): 413–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLawrence M, Huber W, Pages H, et al.: Software for computing and annotating genomic ranges. PLoS Comput Biol. 2013; 9(8): e1003118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosenbloom KR, Armstrong J, Barber GP, et al.: The UCSC Genome Browser database: 2015 update. Nucleic Acids Res. 2015; 43(Database issue): D670–D681. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text\n\nSandve GK, Nekrutenko A, Taylor J, et al.: Ten simple rules for reproducible computational research. PLoS Comput Biol. 2013; 9(10): e1003285. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXie Y: Dynamic Documents with R and knitr. CRC Press. 2013; 216. Reference Source\n\nRstudio. RMarkdown: Dynamic Documents for R. 2014. Reference Source\n\nHester J: Knitr Bootstrap framework, R package. 2015. Reference Source"
}
|
[
{
"id": "8554",
"date": "18 May 2015",
"name": "Timothy J. Triche Jr",
"expertise": [],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nNeeds more figures to demonstrate why a user would choose this tool. For example http://leekgroup.github.io/regionReportSupp/bumphunter-example/index.html (but even better would be to show an example, e.g. ITGB2 exon inclusion/exclusion or multiscale DMRs, where in our hands at least, nothing else short of IGV really does the job, and IGV doesn't do it that well.) The software is a firm foundation but the writeup needs work if it is to be compelling and thus influence readers to try out an unfamiliar tools.My apologies for being harsh, but without figures, an applied paper simply will not be read. I would be less harsh if the underlying work were not compelling enough to command broader interest. A poor writeup will doom the work to obscurity.",
"responses": [
{
"c_id": "1362",
"date": "18 May 2015",
"name": "Jeffrey Leek",
"role": "Author Response",
"response": "Thanks, we will update with more figures and expand the description. This is meant to be a short description of the software but we certainly appreciate the feedback on how to increase users. We will update the draft and respond shortly."
}
]
},
{
"id": "8558",
"date": "11 Jun 2015",
"name": "Karthik Ram",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nReview of regionReport: Interactive reports for region based analysis.This short software tool article describes a new R package, `regionReport`, available from Bioconductor that generates HTML reports which allows users to explore genomic regions and quickly scan quality control information. The reports also provide provenance of code used in the analysis, including detailed session information to facilitate as much reproducibility as possible.The paper/report clearly describes functionality of the tool, potential use cases, and details on installation and operation. Suggestions for improvementGiven that you are describing an HTML application, it would be really helpful to include more screenshots/figures rather than just the one, and the workflow diagram. Most readers are unlikely to install the package immediately (see further comments on 3) and so it would help to make the value proposition clear. I would also suggest annotating these figures highlighting the key parts. Happy to approve the narrative itself after this revision. Given that the package primarily generates html based reports, it would be of great value to have these files in the `gh-pages` branch of a GitHub repo, such that reports could be automatically made available under `https://USERNAME.github.io/repo/file.html` much like the page that describes the supplementary material. One way to easily enable this would be to use the functionality in the `git2r` package (disclosure: I am a coauthor on the package) to programmatically create a new branch (if it doesn't already exist), generate the report, then add those files and push to GitHub (assuming the same folder in under git revision control). Obviously I am not expecting the authors to add this suggestion to the current version of the package, but as something to consider for future versions. Reduce the number of dependencies. `locfdr` is no longer on CRAN. On a slightly slower than normal connection (currently on travel) it took a fairly long time to track down and install all the dependencies. I'd recommend moving non-essential dependencies to suggests and using something like this to selectively install packages as needed using `requireNamespace(pkg, quietly = TRUE)`. It was disappointing to go down a rabbit hole of dependencies and still not be able to install and run examples. However, I found the report examples posted online (here: http://leekgroup.github.io/regionReportSupp/bumphunter-example/index.html and here: http://leekgroup.github.io/regionReportSupp/DiffBind-example/index.html) extremely useful. The use of Twitter bootstrap also adds a layer of a familiarity that I found extremely useful. It would be nice to have the package generate a direct link to the bib file under the bibliography.",
"responses": []
},
{
"id": "8559",
"date": "22 Jun 2015",
"name": "David Robinson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present regionReport, an R package to produce interactive HTML reports from a genomic-region based analysis, such as those produced by derfinder, bumphunter or DiffBind. The report shows quality control summaries, interactive tables of the most significant regions, and information on how regions lie in exons, introns, and intergenic regions. Both novice and expert genomicists are sure to gain much from this, and the paper clearly and concisely explains the software and its use, while providing useful examples and instructions.The idea of producing common reports for a Bioconductor object is ingenious, and will hopefully inspire packages for other types of biological data. One of the great strengths of the package is the reproducibility practices it follows. For example, the section at the end of the produced report that shows reproducibility information, such as the original command, the session info, and the amount of time the report took to generate, is a great idea. (Indeed, the option to add sessionInfo() and timers could probably be baked into rmarkdown, or a thin wrapper thereof). Another strength is the use of modern knitr templates, such as expandable tables. Scientists who want to develop automated reports should use this package as a guide.Overall my concerns are minor, and mostly concern the package rather than the paper, some of which I attempt to address in a GitHub pull request.In pull requestIf the renderReport function leaves early (for example, if it is interrupted by the user hitting Stop) it strands the user's R session in a working directory. Using the on.exit function, as described here, lets R return to the original directory instead. The options for customization of the report are limited, by the customCode argument, to chunks between the main text and the reproducibility section. Genomicists may wish to take advantage of these reports while customizing some of their outputs. (For example, the authors of region-finding packages may wish to wrap renderReport with a customized template for their own objects). I've added a template argument in my pull request, and go over another suggestion below.Not in pull requestThe `template` argument is a start towards greater customization, but a further improvement would be to allow the user to provide a list of customized internal chunks (for example, density-pvalue). As it is now, these are constructed in the renderReport function and cannot be altered without rewriting the entire function. This suggests finding a way to abstract them, such as bringing them in from a separate file, would be useful. As one example of an important customization I'd make: the reports show density plots of p-values and q-values, but in my experience genomicists are more accustomed to histograms (especially since bumps in density plots may be misleading, while histograms can get a better sense of which bumps are meaningful). I understand if the authors wish to keep it as a density plot, but if so I would appreciate a way to change it for my own use.Minor issuesThe use of \"smart quotes\" in code within the PDF, such as source (“http://bioconductor.org/biocLite.R”), make it inconvenient to copy and paste them into an R terminal. If there's any way this could be remedied by the author or editors, it should.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/4-105
|
https://f1000research.com/articles/5-1541/v1
|
29 Jun 16
|
{
"type": "Review",
"title": "Recent Insights into Cell Surface Heparan Sulphate Proteoglycans and Cancer",
"authors": [
"John R Couchman",
"Hinke Multhaupt",
"Ralph D. Sanderson",
"Hinke Multhaupt",
"Ralph D. Sanderson"
],
"abstract": "A small group of cell surface receptors are proteoglycans, possessing a core protein with one or more covalently attached glycosaminoglycan chains. They are virtually ubiquitous and their chains are major sites at which protein ligands of many types interact. These proteoglycans can signal and regulate important cell processes, such as adhesion, migration, proliferation, and differentiation. Since many protein ligands, such as growth factors, morphogens, and cytokines, are also implicated in tumour progression, it is increasingly apparent that cell surface proteoglycans impact tumour cell behaviour. Here, we review some recent advances, emphasising that many tumour-related functions of proteoglycans are revealed only after their modification in processes subsequent to synthesis and export to the cell surface. These include enzymes that modify heparan sulphate structure, recycling of whole or fragmented proteoglycans into exosomes that can be paracrine effectors or biomarkers, and lateral interactions between some proteoglycans and calcium channels that impact the actin cytoskeleton.",
"keywords": [
"Heparan sulphate",
"heparanase",
"exosomes",
"Cell surface proteoglycans"
],
"content": "Introduction\n\nProteoglycans are present in all cellular and tissue compartments. Moreover, in mammals they are expressed by virtually all cells. By definition, proteoglycans consist of a core protein to which one or more glycosaminoglycan chains are covalently attached. While the number of proteoglycan core proteins in the mammalian genome is not large, their form and functions are highly variable. Aggrecan, a major constituent of cartilage matrix, for example, may have >100 chondroitin sulphate chains, which are key to its function in the maintenance of a hydrated, compression-resisting matrix1,2. Decorin, on the other hand, with roles in collagen fibril formation and regulation of innate immunity, has only one chondroitin or dermatan sulphate chain3. Not surprisingly, since proteoglycans can be intracellular, cell surface, or extracellular matrix components, they are increasingly studied in the context of tumour growth, the tumour and stem cell niche, and invasion, metastasis, and tumour-host interactions4–9.\n\nOn the surfaces of most mammalian cells are representatives of two major families of heparan sulphate proteoglycans (HSPGs), the glypicans and syndecans5,10–12. The former are linked to the membrane through a glycosylphosphatidylinositol anchor, while the syndecans are transmembrane, with a highly conserved short cytoplasmic domain. Usually the core proteins carry two to five heparan sulphate chains, but syndecans may sometimes also, or alternately, carry chondroitin or dermatan sulphate chains5. The synthesis of heparan sulphate chains is a complex Golgi apparatus-localised process; while all of the transferases and other modifying enzymes involved in their synthesis are known, their regulation is not13. The importance of heparan sulphate synthesis lies in the fact that this glycosaminoglycan has an ability to interact with a wide array of binding partners that include cytokines, chemokines, growth factors, extracellular matrix macromolecules, enzymes, and lipoproteins14,15. Heparan sulphate chains have regions of high modification (i.e. high levels of sulphation) interspersed with regions of low, or no, sulphation15. This most complex of all post-translational modifications is under scrutiny, since most protein binding partners of heparan sulphate engage with highly sulphated domains14,16, so the control of its synthesis and how this may change with transformation are important issues. Moreover, mature heparan sulphate chains can be further modified by a single mammalian heparanase enzyme and by two sulphatases that selectively remove the sulphates of some glucosamine residues17–19. Heparan sulphate editing is now a topic of great interest in tumour biology and some recent developments are summarised below.\n\nFor many years, it was assumed that cell surface HSPGs had few independent functions but were mostly acting in cis as co-receptors with other receptors, e.g., tyrosine kinase growth factor receptors and integrins5,11,12,20. The notion was that the heparan sulphate chains provided binding sites for ligands that could then be concentrated for high-affinity receptor binding and subsequent signalling. It now seems clear that there are more intricate interactions at the cell surface that involve independent roles for the cell surface HSPGs. Some of the latest insights into cell surface HSPG functions with relevance to tumour biology are briefly reviewed here. Recent information on the roles of other classes of extracellular matrix proteoglycans in cancer can be found elsewhere3,4,7,9,21.\n\n\nHeparan sulphate editing: regulatory events in tumour progression\n\nThere is abundant evidence that heparan sulphates, owing to their diversity in structure and location, play important roles in regulating the growth and progression of cancer. Much of this regulation occurs via the ability of heparan sulphate to fine-tune molecular interactions that regulate cell behaviour22. Over the last decade, it has become increasingly apparent that enzymes can edit heparan sulphate structure, thereby precisely modulating its function and regulating cell behaviour. These enzymes include the endoglucuronidase heparanase, which cleaves and shortens heparan sulphate chains of proteoglycans that as a consequence possess new non-reducing termini, and the extracellular sulphatases Sulf-1 and -2 that selectively remove 6-O sulphates. Both of these enzyme activities are proving to be powerful regulators of tumour behaviour.\n\nHeparanase is associated with aggressive tumour behaviour including enhanced growth, angiogenesis, and metastasis. Although a number of studies in many tumour types have supported these conclusions, a unifying mechanistic explanation of precisely how heparanase promotes angiogenesis and metastasis was lacking until recently. In a paper just published in Oncogenesis, Jung et al. demonstrate that heparanase-mediated trimming of syndecan-1 heparan sulphate chains and upregulation of matrix metalloproteinase-9 (MMP-9) expression results in enhanced shedding of syndecan-1 from the cell surface. Shedding exposes a juxtamembrane site on the syndecan-1 core protein that binds to both very late antigen-4 (VLA-4 [integrin α4β1]) and vascular endothelial growth factor receptor-2 (VEGFR2). This coupling of VLA-4 to VEGFR2 activates the latter, thereby initiating downstream signalling that displaces the cytoskeletal adaptor protein paxillin from VLA-4, in turn facilitating the activation of Rac GTPase and polarised cell migration23. This mechanism is in play on both endothelial cells and tumour cells and demonstrates how heparanase, in concert with syndecan-1, drives angiogenesis, tumour cell invasion, and subsequent metastasis.\n\nEvidence is also emerging that heparanase plays a key role in promoting chemoresistance. In breast cancer cell lines expressing a high level of heparanase, inhibition of the enzyme sensitised the cells to killing by lapatinib24. Elevated heparanase expression by myeloma cells enhances their resistance to both bortezomib and melphalan and this resistance is reversed in vivo when mice are treated with the heparanase inhibitor Roneparstat25. Furthermore, heparanase was shown to be present at a high level on tumour cells that survive extensive chemotherapy in myeloma patients, lending further support to the notion that heparanase promotes resistance to therapy25. Together, these findings raise the exciting possibility that the efficacy of anti-cancer drugs may be enhanced when combined with the use of heparanase inhibitors. This is of particular interest, as there are currently four anti-heparanase drugs in clinical trials in cancer patients19. These drugs are all heparin mimetics that are thought to inhibit heparanase activity by blocking the enzyme’s active site. However, recent solving of the crystal structure of heparanase provides an opportunity for the discovery of small molecule inhibitors of enzyme activity that should exhibit improved specificity over the heparin mimetics26. Heparanase-neutralising antibodies have also recently shown promise in attenuating the growth and metastasis of lymphoma and myeloma tumours in mice27.\n\nWhile heparanase may have important roles in supporting tumour angiogenesis, it is important to recognise that it is not the only mechanism. Many angiogenesis-promoting growth factors, such as VEGF, fibroblast growth factors (FGFs), cytokines, and chemokines, have high affinity for heparan sulphate. It is therefore likely that vascular remodelling is a consequence of multiple interactions involving cell surface HSPGs14,28–30.\n\nAlthough it is generally agreed that the function of Sulf-1 and -2 is to selectively remove 6-O sulphates from heparan sulphate chains, the impact of these two extracellular sulphatases on tumour growth and progression remains controversial. By altering the composition of heparan sulphates, the Sulfs regulate the signalling capacity of heparin-binding growth factors such as Wnts, FGF, EGF, and VEGF, among others19. Predictably, this has important consequences for tumour behaviour. What is surprising is that despite their seemingly identical function, there are data to support the conclusion that Sulf-1 suppresses tumour growth while Sulf-2 promotes tumour growth31,32. However, such a generalisation appears to be misleading because there is evidence that in some instances Sulf-1 promotes, while Sulf-2 inhibits, tumour growth. Together, these findings strongly suggest that there are factors beyond the catalytic activity of the Sulfs that determine their ultimate impact on tumour behaviour31,33,34 (Figure 1). Such factors may be related to spatial or temporal expression of the Sulfs, variations in their specificity for the heparan sulphate substrate, or differing abilities of the Sulfs to diffuse through the tumour microenvironment. Moreover, there is evidence for non-catalytic properties of Sulfs that lead to alterations in heparan sulphate synthesis through changes in sulphotransferase expression33 or upregulation of glypican-3 core protein, which is relevant to hepatocellular carcinoma34.\n\nCell surface heparan sulphate proteoglycans can interact with multiple ligands through their glycosaminoglycan chains. In addition, they can be modified by heparanase and sulphatases, leading to altered ligand binding. Endocytosis, trafficking, and processing can lead to the release of exosomes bearing modified proteoglycans. These can interact with fibronectin in the extracellular environment and ultimately be bound and internalised by recipient cells. This signalling at a distance may be important in the regulation of tumour cell behaviour.\n\n\nSignalling at a distance through exosomes\n\nIn 2012, the first of several papers was published suggesting that syndecans were cell surface receptors important in exosome formation35. For this, the most C-terminal region of the syndecan cytoplasmic domain interacting with PDZ domain proteins was required. The cytoplasmic scaffolding protein syntenin (also known as melanoma differentiation-associated gene 9; MDA-9) binds to all syndecans through one of its two PDZ domains36,37, and this was shown to be important for the endosomal and trafficking events that lead to exosome formation38. The other PDZ domain of syntenin had high affinity for the membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2). Syntenin also interacts through its C-terminal domain with Bro1/ALG-2-interacting protein (ALIX39), a central player in exosome formation. In turn, ALIX links to a multiprotein endosomal sorting complex required for transport (ESCRT), with additional roles for the GTPase Arf6 and phospholipase D240. Exosomes are now recognised as important signalling vesicles, containing a number of proteins, lipids, and even nucleic acids such as RNAs and miRNAs. They are produced by most cells, including tumour cells, and interest in them from the tumour perspective focuses on whether they can be detectable biomarkers in fluids and their potential roles in regulating the tumour environment (Figure 1). Moreover, syntenin (MDA-9) was first identified in the context of melanoma but is upregulated in many tumours where experiments have shown that it supports cell migration or invasion37,41. It has many binding partners beyond syndecans, including the tetraspanin CD63, an exosome marker42, but what controls the selectivity of syntenin to interact with many different cell surface molecules is currently unclear. However, it has been suggested that this protein is a potential tumour target43.\n\nInterestingly, similar to their roles in regulating tumour angiogenesis and metastasis, heparanase and syndecans also work together in regulating exosome secretion by tumour cells. Enhanced heparanase expression in tumour cells stimulates exosome biogenesis, alters exosome protein composition, and enhances the ability of exosomes to promote tumour cell spreading and endothelial cell migration44. In this instance, heparan sulphate chains of syndecans are essential for exosome formation within endosomal compartments, and trimming of heparan sulphate by heparanase activates the formation of an endosomal complex containing syndecan coupled to syntenin and ALIX35,45. This complex promotes endosomal membrane budding and drives exosome biogenesis. Following their secretion, exosomes exert their biological activity by docking with recipient cells and delivering cargo that can alter recipient cell behaviour. In this context, the heparan sulphate present on syndecan, which remains on the exosome surface following the biogenesis process, can interact with fibronectin via its Hep-II heparin-binding domain46. The fibronectin-coated exosomes subsequently dock by binding to the heparan sulphate of proteoglycans present on the recipient cell surface. At least in some cases, the heparan sulphate present on recipient cells can also act as an internalising receptor, thus facilitating the uptake of exosomes and subsequent delivery of exosome cargo within the cell47 (Figure 1).\n\nSyndecans are not the only proteoglycans with potential importance to exosomes. In 2015, a very interesting report documented that circulating exosomes containing glypican-1 could potentially identify patients with pancreatic ductal adenocarcinoma, even at early stages of tumour development48. Whether the heparan sulphate chains were present and carrying important growth factors, cytokines, or chemokines remains speculative, but once more the connection between cell surface HSPGs and cancer is apparent.\n\n\nSyndecans, cytoskeleton, adhesion, and migration\n\nThe four mammalian syndecans all interact with the actin cytoskeleton5. Much research has been devoted to understanding this relationship, and many reports have provided evidence that they contribute to microfilament organisation in adhesion and migration. Perhaps the best example in this regard is syndecan-4. It promotes the assembly of focal adhesions, junctions that form in response to cell adhesion to the extracellular matrix. They are integrin-dependent organelles, but the mechanism by which syndecan influences the process has taken many years to unravel. Key to syndecan-4’s role are interactions with both the actin-associated protein α-actinin49–51 and protein kinase Cα, through which there are multiple potential pathways involving Rho family GTPases to the cytoskeleton52,53. The roles of RhoA, Rac, and cdc42 are well known in this regard54,55. Analysis of fibroblasts derived from syndecan-4 null mice show clear differences in microfilament organisation, with much reduced focal adhesions and stress fibres51,56,57, for which RhoGTPase activities seem not to provide the whole explanation. Recent analysis has now shown that this altered adhesion phenotype of S4KO cells relates to calcium channels of the TRPC (transient receptor potential canonical) family. Indeed, elimination of the TRPC7 channel (itself a focal adhesion component) reverts the S4KO cells to wild-type in terms of adhesion, cytoskeleton, and junction formation58. This was accompanied by reductions in cytosolic calcium that were shown to be increased in the null cells compared to matching wild-type cells. Further work with epithelial cells and, moreover, genetic experiments with Caenorhabditis elegans (which possesses a single syndecan) show that this regulation of TRPC type channels by syndecans may be a highly conserved and important role for this proteoglycan family58.\n\nThe work with syndecans and channels has so far not embraced tumour cells. Since calcium is a potent regulator of the actin cytoskeleton, it may now be attractive to re-examine some of the previous observations on HSPGs and tumour cells. The literature is replete with studies showing that syndecans are often mis-expressed in solid tumours and in some cases relate to prognosis59–62. A good example is breast cancer, where high levels of syndecan-1 expression, particularly in the tumour stroma, are an indicator of poor prognosis63,64. In other studies, syndecan-2 upregulation has been shown to alter the adhesion and invasiveness of MDA-MB231 breast carcinoma cells and colon carcinoma cells65,66. The difficulty with many studies is understanding whether syndecan expression merely correlates with or is functionally related to tumour progression. In some cases, however, the situation is clearer. A wealth of evidence now suggests that syndecan-1 expression in myeloma is related directly to disease severity and progression67,68. Moreover, it is not only syndecans that may influence tumour progression. Evidence has accumulated rapidly over the past few years showing a relationship between glypican-3 expression and the progression of hepatocellular carcinoma69–71. This HSPG is expressed in foetal liver, but levels subside in postnatal life72,73. However, in a large majority of cases, glypican-3 is re-expressed in hepatocellular carcinoma72,74. The excitement about this HSPG revolves around the possibility that it may serve as a prognostic marker, but also a target for immunotherapy71. Early clinical trials have been reported, but clearly there is a long way to go. On a molecular level, it has been suggested that glypican-3 can bind both Wnt and Frizzled, the signalling receptor for Wnts, through its heparan sulphate chains70,71. However, the situation is complex, since glypican-3 in normal tissue may be a growth inhibitor. Rare core protein mutations giving rise to the Simpson-Golabi-Behmel syndrome are characterised by overgrowth and many dysmorphisms in patients and a corresponding murine model75. In hepatocellular carcinoma, however, there is also upregulation of Sulf-2. It now appears that selective removal of 6-O-sulphate residues from the glypican’s heparan sulphate chains leads to Wnt activation, possibly through its enhanced mobility, leading to Frizzled binding and signalling76. It is also possible that the heparan sulphate chains may bind hepatocyte growth factor and members of the FGF family77,78.\n\n\nConclusions\n\nRecent developments have highlighted that both the heparan sulphate chains and the core proteins of cell surface HSPGs are highly and functionally relevant to tumour progression. Moreover, the increasingly recognised importance of the tumour cell niche79,80, which is rich in proteoglycans, and the emerging roles of proteoglycans in stem cell differentiation6,81 are areas for future development. Moreover, it is not only HSPGs that present as targets in tumours. The chondroitin sulphate proteoglycan 4 (also known as NG2) is recognised as a cell surface marker of pericytes in the vasculature but is also present more widely, for example on neuronal and oligodendrocyte precursors82. It is also an emerging target for immunotherapy in a variety of tumour types, including melanoma, triple negative breast cancer, glioblastoma, mesothelioma, and sarcomas83,84.\n\nThe potential for cell surface proteoglycans to be targets for intervention are complicated by their multiple roles and ubiquity. It is perhaps likely that tumour cells, stromal/other host tissue, and the immune system utilise these proteoglycans and their downstream signalling in specific ways to regulate behaviour. Targeting will require detailed understanding, and therefore we can predict that new insights into the functions of proteoglycans will impact tumour biology for many years to come.",
"appendix": "Author contributions\n\n\n\nAll three authors contributed to the writing and editing of this review. HAB was responsible for the figure.\n\n\nCompeting interests\n\n\n\nRDS is on the Scientific Advisory Board of Sigma-tau Research S.A.\n\n\nGrant information\n\nJRC and HAB were supported by the Department of Biomedical Sciences and the Biotech Research & Innovation Center at the University of Copenhagen. RDS is supported by National Institutes of Health grant CA138340.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAspberg A: The different roles of aggrecan interaction domains. J Histochem Cytochem. 2012; 60(12): 987–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeinegård D, Saxne T: The role of the cartilage matrix in osteoarthritis. Nat Rev Rheumatol. 2011; 7(1): 50–6. PubMed Abstract | Publisher Full Text\n\nNeill T, Schaefer L, Iozzo RV: Decoding the Matrix: Instructive Roles of Proteoglycan Receptors. Biochemistry. 2015; 54(30): 4583–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIozzo RV, Sanderson RD: Proteoglycans in cancer biology, tumour microenvironment and angiogenesis. J Cell Mol Med. 2011; 15(5): 1013–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCouchman JR: Transmembrane signaling proteoglycans. Annu Rev Cell Dev Biol. 2010; 26: 89–114. PubMed Abstract | Publisher Full Text\n\nPickford CE, Holley RJ, Rushton G, et al.: Specific glycosaminoglycans modulate neural specification of mouse embryonic stem cells. Stem Cells. 2011; 29(4): 629–40. PubMed Abstract | Publisher Full Text\n\nSchaefer L, Iozzo RV: Small leucine-rich proteoglycans, at the crossroad of cancer growth and inflammation. Curr Opin Genet Dev. 2012; 22(1): 56–7. PubMed Abstract | Publisher Full Text\n\nMulthaupt HA, Leitinger B, Gullberg D, et al.: Extracellular matrix component signaling in cancer. Adv Drug Deliv Rev. 2016; 97: 28–40. PubMed Abstract | Publisher Full Text\n\nTheocharis AD, Gialeli C, Bouris P, et al.: Cell-matrix interactions: focus on proteoglycan-proteinase interplay and pharmacological targeting in cancer. FEBS J. 2014; 281(22): 5023–42. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nFilmus J, Capurro M, Rast J: Glypicans. Genome Biol. 2008; 9(5): 224. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan MR, Humphries MJ, Bass MD: Synergistic control of cell adhesion by integrins and syndecans. Nat Rev Mol Cell Biol. 2007; 8(12): 957–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCouchman JR, Gopal S, Lim HC, et al.: Syndecans: from peripheral coreceptors to mainstream regulators of cell behaviour. Int J Exp Pathol. 2015; 96(1): 1–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMulthaupt HA, Couchman JR: Heparan sulfate biosynthesis: methods for investigation of the heparanosome. J Histochem Cytochem. 2012; 60(12): 908–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu D, Esko JD: Demystifying heparan sulfate-protein interactions. Annu Rev Biochem. 2014; 83: 129–57. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGallagher J: Fell-Muir Lecture: Heparan sulphate and the art of cell regulation: a polymer chain conducts the protein orchestra. Int J Exp Pathol. 2015; 96(4): 203–31. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLindahl U, Li JP: Interactions between heparan sulfate and proteins-design and functional implications. Int Rev Cell Mol Biol. 2009; 276: 105–59. PubMed Abstract | Publisher Full Text\n\nLamanna WC, Kalus I, Padva M, et al.: The heparanome--the enigma of encoding and decoding heparan sulfate sulfation. J Biotechnol. 2007; 129(2): 290–307. PubMed Abstract | Publisher Full Text\n\nRosen SD, Lemjabbar-Alaoui H: Sulf-2: an extracellular modulator of cell signaling and a cancer target candidate. Expert Opin Ther Targets. 2010; 14(9): 935–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHammond E, Khurana A, Shridhar V, et al.: The Role of Heparanase and Sulfatases in the Modification of Heparan Sulfate Proteoglycans within the Tumor Microenvironment and Opportunities for Novel Cancer Therapeutics. Front Oncol. 2014; 4: 195. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBernfield M, Götte M, Park PW, et al.: Functions of cell surface heparan sulfate proteoglycans. Annu Rev Biochem. 1999; 68: 729–77. PubMed Abstract | Publisher Full Text\n\nPoluzzi C, Iozzo RV, Schaefer L: Endostatin and endorepellin: A common route of action for similar angiostatic cancer avengers. Adv Drug Deliv Rev. 2016; 97: 156–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBishop JR, Schuksz M, Esko JD: Heparan sulphate proteoglycans fine-tune mammalian physiology. Nature. 2007; 446(7139): 1030–7. PubMed Abstract | Publisher Full Text\n\nJung O, Trapp-Stamborski V, Purushothaman A, et al.: Heparanase-induced shedding of syndecan-1/CD138 in myeloma and endothelial cells activates VEGFR2 and an invasive phenotype: prevention by novel synstatins. Oncogenesis. 2016; 5: e202. PubMed Abstract | Publisher Full Text\n\nZhang L, Ngo JA, Wetzel MD, et al.: Heparanase mediates a novel mechanism in lapatinib-resistant brain metastatic breast cancer. Neoplasia. 2015; 17(1): 101–13. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRamani VC, Zhan F, He J, et al.: Targeting heparanase overcomes chemoresistance and diminishes relapse in myeloma. Oncotarget. 2016; 7(2): 1598–607. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu L, Viola CM, Brzozowski AM, et al.: Structural characterization of human heparanase reveals insights into substrate recognition. Nat Struct Mol Biol. 2015; 22(12): 1016–22. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWeissmann M, Arvatz G, Horowitz N, et al.: Heparanase-neutralizing antibodies attenuate lymphoma tumor growth and metastasis. Proc Natl Acad Sci U S A. 2016; 113(3): 704–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nJakobsson L, Kreuger J, Holmborn K, et al.: Heparan sulfate in trans potentiates VEGFR-mediated angiogenesis. Dev Cell. 2006; 10(5): 625–34. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMatsuo I, Kimura-Yoshida C: Extracellular modulation of Fibroblast Growth Factor signaling through heparan sulfate proteoglycans in mammalian development. Curr Opin Genet Dev. 2013; 23(4): 399–407. PubMed Abstract | Publisher Full Text\n\nPeysselon F, Ricard-Blum S: Heparin-protein interactions: from affinity and kinetics to biological roles. Application to an interaction network regulating angiogenesis. Matrix Biol. 2014; 35: 73–81. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nVivès RR, Seffouh A, Lortat-Jacob H: Post-Synthetic Regulation of HS Structure: The Yin and Yang of the Sulfs in Cancer. Front Oncol. 2014; 3: 331. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYang JD, Sun Z, Hu C, et al.: Sulfatase 1 and sulfatase 2 in hepatocellular carcinoma: associated signaling pathways, tumor phenotypes, and survival. Genes Chromosomes Cancer. 2011; 50(2): 122–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLamanna WC, Frese MA, Balleininger M, et al.: Sulf loss influences N-, 2-O-, and 6-O-sulfation of multiple heparan sulfate proteoglycans and modulates fibroblast growth factor signaling. J Biol Chem. 2008; 283(41): 27724–35. PubMed Abstract | Publisher Full Text\n\nLai JP, Sandhu DS, Yu C, et al.: Sulfatase 2 up-regulates glypican 3, promotes fibroblast growth factor signaling, and decreases survival in hepatocellular carcinoma. Hepatology. 2008; 47(4): 1211–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaietti MF, Zhang Z, Mortier E, et al.: Syndecan-syntenin-ALIX regulates the biogenesis of exosomes. Nat Cell Biol. 2012; 14(7): 677–85. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGrootjans JJ, Zimmermann P, Reekmans G, et al.: Syntenin, a PDZ protein that binds syndecan cytoplasmic domains. Proc Natl Acad Sci U S A. 1997; 94(25): 13683–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDas SK, Bhutia SK, Kegelman TP, et al.: MDA-9/syntenin: a positive gatekeeper of melanoma metastasis. Front Biosci (Landmark Ed). 2012; 17: 1–15. PubMed Abstract | Publisher Full Text\n\nGhossoub R, Lembo F, Rubio A, et al.: Syntenin-ALIX exosome biogenesis and budding into multivesicular bodies are controlled by ARF6 and PLD2. Nat Commun. 2014; 5: 3477. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nThéry C, Boussac M, Véron P, et al.: Proteomic analysis of dendritic cell-derived exosomes: a secreted subcellular compartment distinct from apoptotic vesicles. J Immunol. 2001; 166(12): 7309–18. PubMed Abstract | Publisher Full Text\n\nFriand V, David G, Zimmermann P: Syntenin and syndecan in the biogenesis of exosomes. Biol Cell. 2015; 107(10): 331–41. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKashyap R, Roucourt B, Lembo F, et al.: Syntenin controls migration, growth, proliferation, and cell cycle progression in cancer cells. Front Pharmacol. 2015; 6: 241. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nEscola JM, Kleijmeer MJ, Stoorvogel W, et al.: Selective enrichment of tetraspan proteins on the internal vesicles of multivesicular endosomes and on exosomes secreted by human B-lymphocytes. J Biol Chem. 1998; 273(32): 20121–7. PubMed Abstract | Publisher Full Text\n\nKegelman TP, Das SK, Emdad L, et al.: Targeting tumor invasion: the roles of MDA-9/Syntenin. Expert Opin Ther Targets. 2015; 19(1): 97–112. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nThompson CA, Purushothaman A, Ramani VC, et al.: Heparanase regulates secretion, composition, and function of tumor cell-derived exosomes. J Biol Chem. 2013; 288(14): 10093–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoucourt B, Meeussen S, Bao J, et al.: Heparanase activates the syndecan-syntenin-ALIX exosome pathway. Cell Res. 2015; 25(4): 412–28. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPurushothaman A, Bandari SK, Liu J, et al.: Fibronectin on the Surface of Myeloma Cell-derived Exosomes Mediates Exosome-Cell Interactions. J Biol Chem. 2016; 291(4): 1652–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChristianson HC, Svensson KJ, van Kuppevelt TH, et al.: Cancer cell exosomes depend on cell-surface heparan sulfate proteoglycans for their internalization and functional activity. Proc Natl Acad Sci U S A. 2013; 110(43): 17380–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelo SA, Luecke LB, Kahlert C, et al.: Glypican-1 identifies cancer exosomes and detects early pancreatic cancer. Nature. 2015; 523(7559): 177–82. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGreene DK, Tumova S, Couchman JR, et al.: Syndecan-4 associates with alpha-actinin. J Biol Chem. 2003; 278(9): 7617–23. PubMed Abstract | Publisher Full Text\n\nChoi Y, Kim S, Lee J, et al.: The oligomeric status of syndecan-4 regulates syndecan-4 interaction with alpha-actinin. Eur J Cell Biol. 2008; 87(10): 807–15. PubMed Abstract | Publisher Full Text\n\nOkina E, Grossi A, Gopal S, et al.: Alpha-actinin interactions with syndecan-4 are integral to fibroblast-matrix adhesion and regulate cytoskeletal architecture. Int J Biochem Cell Biol. 2012; 44(12): 2161–74. PubMed Abstract | Publisher Full Text\n\nDovas A, Yoneda A, Couchman JR: PKCbeta-dependent activation of RhoA by syndecan-4 during focal adhesion formation. J Cell Sci. 2006; 119(Pt 13): 2837–46. PubMed Abstract | Publisher Full Text\n\nBass MD, Roach KA, Morgan MR, et al.: Syndecan-4-dependent Rac1 regulation determines directional migration in response to the extracellular matrix. J Cell Biol. 2007; 177(3): 527–38. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHall A: Rho family GTPases. Biochem Soc Trans. 2012; 40(6): 1378–82. PubMed Abstract | Publisher Full Text\n\nLi H, Peyrollier K, Kilic G, et al.: Rho GTPases and cancer. Biofactors. 2014; 40(2): 226–35. PubMed Abstract | Publisher Full Text\n\nGopal S, Bober A, Whiteford JR, et al.: Heparan sulfate chain valency controls syndecan-4 function in cell adhesion. J Biol Chem. 2010; 285(19): 14247–58. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMostafavi-Pour Z, Askari JA, Parkinson SJ, et al.: Integrin-specific signaling pathways controlling focal adhesion formation and cell migration. J Cell Biol. 2003; 161(1): 155–67. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGopal S, Søgaard P, Multhaupt HA, et al.: Transmembrane proteoglycans control stretch-activated channels to set cytosolic calcium levels. J Cell Biol. 2015; 210(7): 1199–211. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReijmers RM, Spaargaren M, Pals ST: Heparan sulfate proteoglycans in the control of B cell development and the pathogenesis of multiple myeloma. FEBS J. 2013; 280(10): 2180–93. PubMed Abstract | Publisher Full Text\n\nRapraeger AC: Synstatin: a selective inhibitor of the syndecan-1-coupled IGF1R-αvβ3 integrin complex in tumorigenesis and angiogenesis. FEBS J. 2013; 280(10): 2207–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamani VC, Purushothaman A, Stewart MD, et al.: The heparanase/syndecan-1 axis in cancer: mechanisms and therapies. FEBS J. 2013; 280(10): 2294–306. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTheocharis AD, Skandalis SS, Neill T, et al.: Insights into the key roles of proteoglycans in breast cancer biology and translational medicine. Biochim Biophys Acta. 2015; 1855(2): 276–300. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarbareschi M, Maisonneuve P, Aldovini D, et al.: High syndecan-1 expression in breast carcinoma is related to an aggressive phenotype and to poorer prognosis. Cancer. 2003; 98(3): 474–83. PubMed Abstract | Publisher Full Text\n\nLeivonen M, Lundin J, Nordling S, et al.: Prognostic value of syndecan-1 expression in breast cancer. Oncology. 2004; 67(1): 11–8. PubMed Abstract | Publisher Full Text\n\nLim HC, Multhaupt HA, Couchman JR: Cell surface heparan sulfate proteoglycans control adhesion and invasion of breast carcinoma cells. Mol Cancer. 2015; 14: 15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRyu HY, Lee J, Yang S, et al.: Syndecan-2 functions as a docking receptor for pro-matrix metalloproteinase-7 in human colon cancer cells. J Biol Chem. 2009; 284(51): 35692–701. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhotskaya YB, Dai Y, Ritchie JP, et al.: Syndecan-1 is required for robust growth, vascularization, and metastasis of myeloma tumors in vivo. J Biol Chem. 2009; 284(38): 26085–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamani VC, Sanderson RD: Chemotherapy stimulates syndecan-1 shedding: a potentially negative effect of treatment that may promote tumor relapse. Matrix Biol. 2014; 35: 215–22. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFilmus J, Capurro M: Glypican-3: a marker and a therapeutic target in hepatocellular carcinoma. FEBS J. 2013; 280(10): 2471–6. PubMed Abstract | Publisher Full Text\n\nCapurro M, Martin T, Shi W, et al.: Glypican-3 binds to Frizzled and plays a direct role in the stimulation of canonical Wnt signaling. J Cell Sci. 2014; 127(Pt 7): 1565–75. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nHaruyama Y, Kataoka H: Glypican-3 is a prognostic factor and an immunotherapeutic target in hepatocellular carcinoma. World J Gastroenterol. 2016; 22(1): 275–83. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nCapurro M, Wanless IR, Sherman M, et al.: Glypican-3: a novel serum and histochemical marker for hepatocellular carcinoma. Gastroenterology. 2003; 125(1): 89–97. PubMed Abstract | Publisher Full Text\n\nIglesias BV, Centeno G, Pascuccelli H, et al.: Expression pattern of glypican-3 (GPC3) during human embryonic and fetal development. Histol Histopathol. 2008; 23(11): 1333–40. PubMed Abstract\n\nYamauchi N, Watanabe A, Hishinuma M, et al.: The glypican 3 oncofetal protein is a promising diagnostic marker for hepatocellular carcinoma. Mod Pathol. 2005; 18(12): 1591–8. PubMed Abstract | Publisher Full Text\n\nPilia G, Hughes-Benzie RM, MacKenzie A, et al.: Mutations in GPC3, a glypican gene, cause the Simpson-Golabi-Behmel overgrowth syndrome. Nat Genet. 1996; 12(3): 241–7. PubMed Abstract | Publisher Full Text\n\nLai JP, Oseini AM, Moser CD, et al.: The oncogenic effect of sulfatase 2 in human hepatocellular carcinoma is mediated in part by glypican 3-dependent Wnt activation. Hepatology. 2010; 52(5): 1680–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZittermann SI, Capurro MI, Shi W, et al.: Soluble glypican 3 inhibits the growth of hepatocellular carcinoma in vitro and in vivo. Int J Cancer. 2010; 126(6): 1291–301. PubMed Abstract | Publisher Full Text\n\nGao W, Kim H, Ho M: Human Monoclonal Antibody Targeting the Heparan Sulfate Chains of Glypican-3 Inhibits HGF-Mediated Migration and Motility of Hepatocellular Carcinoma Cells. PLoS One. 2015; 10(9): e0137664. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nVenning FA, Wullkopf L, Erler JT: Targeting ECM Disrupts Cancer Progression. Front Oncol. 2015; 5: 224. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nStewart MD, Ramani VC, Sanderson RD: Shed syndecan-1 translocates to the nucleus of cells delivering growth factors and inhibiting histone acetylation: a novel mechanism of tumor-host cross-talk. J Biol Chem. 2015; 290(2): 941–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOikari LE, Okolicsanyi RK, Qin A, et al.: Cell surface heparan sulfate proteoglycans as novel markers of human neural stem cell fate determination. Stem Cell Res. 2016; 16(1): 92–104. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nArmulik A, Genové G, Betsholtz C: Pericytes: developmental, physiological, and pathological perspectives, problems, and promises. Dev Cell. 2011; 21(2): 193–215. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBeard RE, Zheng Z, Lagisetty KH, et al.: Multiple chimeric antigen receptors successfully target chondroitin sulfate proteoglycan 4 in several different cancer histologies and cancer stem cells. J Immunother Cancer. 2014; 2: 25. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nCampoli M, Ferrone S, Wang X: Functional and clinical relevance of chondroitin sulfate proteoglycan 4. Adv Cancer Res. 2010; 109: 73–121. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14680",
"date": "29 Jun 2016",
"name": "Jeremy E Turnbull",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14676",
"date": "29 Jun 2016",
"name": "Marion Kusche-Gullberg",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14674",
"date": "29 Jun 2016",
"name": "Jeffrey Esko",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1541
|
https://f1000research.com/articles/5-1533/v1
|
28 Jun 16
|
{
"type": "Review",
"title": "Production of Basal Bodies in bulk for dense multicilia formation",
"authors": [
"Xiumin Yan",
"Huijie Zhao",
"Xueliang Zhu",
"Xiumin Yan",
"Huijie Zhao"
],
"abstract": "Centriole number is normally under tight control and is directly linked to ciliogenesis. In cells that use centrosomes as mitotic spindle poles, one pre-existing mother centriole is allowed to duplicate only one daughter centriole per cell cycle. In multiciliated cells, however, many centrioles are generated to serve as basal bodies of the cilia. Although deuterosomes were observed more than 40 years ago using electron microscopy and are believed to produce most of the basal bodies in a mother centriole-independent manner, the underlying molecular mechanisms have remained unknown until recently. From these findings arise more questions and a call for clarifications that will require multidisciplinary efforts.",
"keywords": [
"ciliogenesis",
"centriole assembly",
"Deuterosome",
"deuterosome-dependent"
],
"content": "Introduction\n\nThe centriole is a cylinder-shaped organelle that serves as the core of the centrosome or the basal body of the cilium1–5. Nascent centriole formation usually depends on pre-existing mother centrioles. Normally in one cell cycle each mother centriole produces only one daughter centriole, that is directly adjacent (Figure 1). Such tight control ensures proper mitosis, since only two centrosomes are required to function as the spindle poles. It also guarantees that the centriole number remains constant after cell division (Figure 1).\n\nThe centrosome in a G1 cell contains a pair of mother-daughter centrioles. Upon entering the S phase, each centriole starts to duplicate one daughter centriole so that the centriole number remains constant after mitosis (a). When the cell enters G0, the mother centriole can be transformed into the basal body to support monocilium formation (b). Alternatively, both the mother centriole-dependent (MCD) and deuterosome-dependent (DD) pathways can be activated to generate an abundance of centrioles for dense multicilia formation (c). The scanning electron microscopy images show a primary cilium in the collecting duct of mouse kidney and multicilia of a multiciliated cell in mouse tracheal epithelium, respectively. Centrioles are drawn in blue and their cartwheels in orange.\n\nCiliogenesis occurs at the G0 or G1 stage of the cell cycle (Figure 1)1,3,6. In vertebrates, most cells can possess a primary cilium, which functions as a sensory organ for diverse environmental signals. Mammalian epithelial tissues such as those lining the inner surface of the trachea, the oviduct, and the brain ventricles, however, have abundant multiciliated cells (MCCs) with hundreds of cilia (Figure 1). These multicilia are motile and their beating is critical for mucus clearance, ovum transport, or cerebrospinal fluid circulation7. How then do such cells generate sufficient numbers of basal bodies?\n\nThe mystery was initially uncovered by electron microscopy (EM) on a variety of MCC-containing tissues in the 1960’s and 1970’s. The mother centriole was observed to be surrounded by multiple daughter centrioles in MCCs. Moreover, many granular or ring-shaped EM structures termed deuterosomes (this name will be used in this review), procentriole precursor bodies, dense granules, and generative complexes were also able to initiate procentriole assembly8–12. Importantly, the deuterosomes were estimated to produce most of the basal bodies required. Nevertheless, it is only recently that we have begun to understand the molecular mechanisms involved, which will be the major focus of this review.\n\n\nMother centriole-dependent centriole assembly\n\nTremendous progress has been made toward understanding how a daughter centriole is born in cycling cells. A group of proteins, including Cep152 and Cep63, are specifically located around the proximal side of the mother centriole. In the G1 phase, the polo-like kinase PLK4 binds to Cep152 to form the site of centriole assembly13–17. In the S phase, a cartwheel structure is formed at the PLK4 site, followed by the assembly of the nine sets of microtubule triplets and other components of the daughter centriole. Centriole assembly is completed by the G2 phase and, following mitosis, each daughter cell inherits a mother-daughter pair of centrioles (Figure 1)1–5.\n\nInterestingly, mother centrioles in cycling cells are capable of generating more than one daughter centriole. For instance, overexpression of PLK4 results in multiple PLK4 foci around the mother centriole and overproduction of daughter centrioles18,19. Overexpression of Cep152 or the cartwheel proteins SAS-6 or STIL also has a similar effect20–23. These observations not only indicate that cycling cells execute the one-daughter-centriole-per-mother rule by restricting the levels of several critical proteins but also suggest that MCCs may break this rule by simply upregulating the protein levels. Indeed, when mouse tracheal epithelial cells (MTECs) are induced to form multicilia, they express high levels of these proteins19,24,25. The importance of PLK4 and Cep152 in mother centriole-dependent (MCD) centriole overduplication of MTECs is also verified19.\n\n\nDeuterosome-dependent centriole assembly\n\nThe discovery of an essential deuterosome component, Deup1 (also called Ccdc67), has promoted the understanding of deuterosome-dependent (DD) centriole biogenesis. Strikingly, Deup1 is a paralog of Cep6319. Cep152 binds to both Cep63 and Deup1 to stabilize them and be recruited, respectively, to the mother centriole and the deuterosome19,26. Therefore, if we consider the Cep63-Cep152-containing proximal ring of the mother centriole as a platform, or ‘cradle’, that supports nascent centriole assembly, deuterosomes are analogous cradles, independent of mother centrioles (Figure 2A). In MTECs, deuterosomes appear initially as foci with zero to two associated procentrioles (Figure 2A-B, stage II). Their sizes then enlarge, accompanied by an increase in procentriole numbers (Figure 2A-B, stage III). They are disassembled upon completion of centriole assembly (Figure 2A)19,27. Usually 50–100 deuterosomes can be found in a MTEC, sufficient for the production of hundreds of centrioles (Figure 2B)19. Mouse ependymal cells (MEPCs) displayed a similar centriole amplification process, but their deuterosomes are usually much larger in size and smaller in number (Figure 2C).\n\n(A) Illustration for centriole amplification stages in MTECs19. Centrioles are drawn in blue and their cartwheels in orange. (B) Three-dimensional structured illumination microscopy (3D-SIM) images for MTECs at early stages (II and III) of centriole amplification. MTECs cultured as described previously19 were immunostained for Deup1, Cep63, and Centrin and imaged using a DeltaVision OMX V3 microscopic system (GE Healthcare). The mother centrioles (arrows) and representative deuterosomes (arrowheads) are magnified 2× to show details. (C) 3D-SIM images showing centriole amplification in MEPCs. MEPCs were isolated from neonatal C57BL/6J mice and cultured as described32. The cells were fixed at day three after serum starvation and immunostained for Deup1, Cep152, and Centrin. The stages (II and IV) are defined as in the MTECs. Note that MEPC deuterosomes (C) are usually much larger than those in MTECs (B). (D) SIM images of two large MEPC deuterosomes immunostained for Deup1 and Centrin (top row) or Cep152 and Centrin (bottom row). Their 3D profiles are also shown. Abbreviations: DD, deuterosome dependent; MCD, mother centriole dependent.\n\nThe beauty of such a DD pathway is obvious: cycling cells only need to turn off the DD pathway by shutting down Deup1 expression to avoid the production of extra centrioles. On the other hand, as MCCs are terminally differentiated and no longer able to enter the cell cycle, turning on the DD pathway and upregulating other genes critical for basal body assembly can safely fulfill their demand on large numbers of basal bodies. For instance, the Multicilin-E2F4/5 complex is known to activate the transcription of Deup1, Plk4, Cep152, and many other centriolar protein genes in MCCs28–30. Other proteins such as cyclin O appear to fine-tune the transcription program31.\n\n\nDeuterosome structures and components\n\nDeuterosome size varies remarkably in different tissues and species: for instance, from 100–200 nm (diameter) in rat or mouse MTECs8,19 to more than 500 nm in the mouse oviduct10. Larger deuterosomes are capable of supporting more procentrioles. As deuterosomes look mostly ring shaped in transmission EM, they were proposed to be roughly sphere shaped, capable of assembling centrioles in all directions8,10. Serial ultra-thin sections of MEPCs support this notion32.\n\nThree-dimensional profiling of subdiffraction images from both MTECs and MEPCs, however, suggests that Deup1 and Cep152 are arranged in a ring-shaped configuration in the deuterosome, with the Cep152 signals enwrapping those of Deup1 from outside (Figure 2C)19. Such a configuration is topologically analogous to the mother centriole cradle. Only the ends of the deuterosome appear relatively amorphous. For instance, in large deuterosomes such as those of MEPCs, the Cep152 signals may exhibit several ‘holes’ at each end (Figure 2D). Procentrioles tend to be assembled on the outer wall of the deuterosome but can be found at both ends as well (Figure 2D)19.\n\nWhether there are additional proteins to construct the outer wall, fill the center, or cap the ends of the deuterosome is presently unknown. Ccdc78, a coiled coil domain-containing protein, is reported as a deuterosome-specific protein required for centriole amplification in the Xenopus embryonic epidermis33. Nonetheless, mouse Ccdc78, expressed either endogenously or exogenously, was not detected on Deup1-positive deuterosomes in our hands, raising the possibility that Ccdc78 may be either an amphibian-specific deuterosome component or even not a bona fide one.\n\n\nDeuterosome assembly\n\nHow deuterosome components are packed together to form the supramolecular structure is also an important issue. Fibrous granules (also called fibrogranular material or proliferative elements), clouds of material abundant in 40 to 80 nm granules that coincide with deuterosome formation in MCCs, were proposed to be precursors of the deuterosome8–10. PCM-1, a component of fibrous granules, however, failed to show deuterosome localization34. Its depletion by RNA interference also didn’t impair centriole amplification25. Likewise, neither Deup1 nor Cep152 exhibited obvious fibrous granule-like distributions (Figure 2B-C)19. Since small deuterosomes tend to emerge in bulk and then grow in synchrony and ectopic expression of Deup1 in cycling cells is sufficient to induce the formation of functional deuterosomes (Figure 2B)19, we propose that deuterosomes can be assembled spontaneously (Figure 2A).\n\nInterestingly, a recent publication argues for a totally different mechanism32. Based mainly on studies in MEPCs, a model is proposed in which an unknown mechanism recruits Deup1, Ccdc78, and other cradle proteins to a site in the cradle of the young mother centriole to initiate the assembly of both the deuterosome and the daughter centrioles. The deuterosome-procentriole halo is then released so that the site can begin the next assembly cycle. After the release of the last halo, procentrioles on all the deuterosomes start to elongate and mature. Thus, both the deuterosome formation and the massive centriole biogenesis are MCD processes. Deuterosomes function merely as shuttles to carry the daughter centrioles away from their mother centriole into the cytoplasm32,35.\n\nThis model, despite its uniqueness, still needs further verification. Firstly, it remains to be shown whether this is the sole and universal way of deuterosome generation. Deup1 is capable of mother centriole localization (Figure 2B-C)19. It is thus understandable that some of the protein there may serve as seeds to initiate deuterosome assembly. Since live imaging in the MEPCs suggests that the generation of one halo requires about two hours32, such an efficiency would demand several days to generate the 50–100 deuterosomes in MTECs, while the entire centriole amplification process takes roughly only one day19,32. Thus, both the spontaneous and MCD pathways may contribute. Furthermore, there might be multiple deuterosome nucleation sites on both the young and the old mother centrioles (Figure 2B-C). Secondly, the model is apparently incompatible with the observation that the numbers of deuterosome-associated procentrioles increase over time (Figure 2B-C)19. Even if one or two daughter centrioles could be carried away from the mother centriole by each nascent deuterosome, their subsequent increase in numbers still argues for the existence of de novo DD centriole biogenesis. Finally, what defines the deuterosome nucleation site on the mother centriole and how the cytoplasmic halos can wait until the last one is released are also issues for future clarification.\n\n\nConservation of the deuterosome-dependent pathway\n\nPhylogenetic analysis suggests that Deup1 is originated from a common fish ancestor of the lobe-finned fish and tetrapods in the vertebrate evolution to boost cilia density in MCCs19,36. Accordingly, in contrast to the lobe-finned fish (such as lungfish), MCCs of the ray-finned fish (such as zebrafish), which have no Deup1, contain only sparse cilia37,38. Many invertebrates, however, possess MCCs with dense multicilia39–42. Deuterosome-like ultrastructures have also been reported in some invertebrate species43,44. A comprehensive knowledge of strategies for centriole amplification throughout metazoan evolution will thus require an understanding of the mechanisms for multiciliogenesis in the invertebrate.\n\n\nConclusions and perspectives\n\nThe mechanism of centriole amplification is both exciting and challenging. Because the sizes of centrioles and deuterosomes are below or close to the optical diffraction limit, technical limitations of imaging are a current major bottleneck restraining studies of centriole amplification in MCCs. Although 3D structured illumination microscopy (SIM)45 has proven its power in the past19,32, the development and introduction of super-resolution techniques with higher spatial (especially the z-axis) and temporal resolutions46–49 are expected to greatly facilitate studies in the field. Furthermore, other cutting-edge techniques such as cryo-electron tomography, omics analysis, and computational biology may help to solve issues on the structure, formation, growth, disassembly, and function of the deuterosome as well as the entire mechanism that controls appropriate on-and-off switching of the centriole amplification program.\n\n\nAbbreviations\n\nDD, deuterosome dependent; EM, electron microscopy; MCC, multiciliated cell; MCD, mother centriole dependent; MEPC, mouse ependymal cell; MTEC, mouse tracheal epithelial cell; SIM, structured illumination microscopy.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThe authors are supported by grants from the National Science Foundation of China (31330045) and the Ministry of Science and Technology of China (2012CB945003).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors thank Dr Nathalie Spassky (CNRS, France) for the mouse ependymal cell culture protocol, Drs Steven Brody and Giuliano Callaini for very helpful criticisms and suggestions on the manuscript, Yawen Chen and Shichao Duan for providing the micrographs of multicilia and primary cilium, and Qijun Tang for the illustrations.\n\n\nReferences\n\nNigg EA, Stearns T: The centrosome cycle: Centriole biogenesis, duplication and inherent asymmetries. Nat Cell Biol. 2011; 13(10): 1154–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBornens M: The centrosome in cells and organisms. Science. 2012; 335(6067): 422–6. PubMed Abstract | Publisher Full Text\n\nIshikawa H, Marshall WF: Ciliogenesis: building the cell's antenna. Nat Rev Mol Cell Biol. 2011; 12(4): 222–34. PubMed Abstract | Publisher Full Text\n\nAvidor-Reiss T, Gopalakrishnan J: Building a centriole. Curr Opin Cell Biol. 2013; 25(1): 72–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGönczy P: Towards a molecular architecture of centriole assembly. Nat Rev Mol Cell Biol. 2012; 13(7): 425–35. PubMed Abstract | Publisher Full Text\n\nGoetz SC, Anderson KV: The primary cilium: a signalling centre during vertebrate development. Nat Rev Genet. 2010; 11(5): 331–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrooks ER, Wallingford JB: Multiciliated cells. Curr Biol. 2014; 24(19): R973–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokin SP: Reconstructions of centriole formation and ciliogenesis in mammalian lungs. J Cell Sci. 1968; 3(2): 207–30. PubMed Abstract\n\nAnderson RG, Brenner RM: The formation of basal bodies (centrioles) in the Rhesus monkey oviduct. J Cell Biol. 1971; 50(1): 10–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDirksen ER: Centriole morphogenesis in developing ciliated epithelium of the mouse oviduct. J Cell Biol. 1971; 51(1): 286–302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalnins VI, Porter KR: Centriole replication during ciliogenesis in the chick tracheal epithelium. Z Zellforsch Mikrosk Anat. 1969; 100(1): 1–30. PubMed Abstract | Publisher Full Text\n\nSteinman RM: An electron microscopic study of ciliogenesis in developing epidermis and trachea in the embryo of Xenopus laevis. Am J Anat. 1968; 122(1): 19–55. PubMed Abstract | Publisher Full Text\n\nCizmecioglu O, Arnold M, Bahtz R, et al.: Cep152 acts as a scaffold for recruitment of Plk4 and CPAP to the centrosome. J Cell Biol. 2010; 191(4): 731–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHatch EM, Kulukian A, Holland AJ, et al.: Cep152 interacts with Plk4 and is required for centriole duplication. J Cell Biol. 2010; 191(4): 721–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSonnen KF, Schermelleh L, Leonhardt H, et al.: 3D-structured illumination microscopy provides novel insight into architecture of human centrosomes. Biol Open. 2012; 1(10): 965–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBettencourt-Dias M, Rodrigues-Martins A, Carpenter L, et al.: SAK/PLK4 is required for centriole duplication and flagella development. Curr Biol. 2005; 15(24): 2199–207. PubMed Abstract | Publisher Full Text\n\nHabedanck R, Stierhof YD, Wilkinson CJ, et al.: The Polo kinase Plk4 functions in centriole duplication. Nat Cell Biol. 2005; 7(11): 1140–6. PubMed Abstract | Publisher Full Text\n\nKleylein-Sohn J, Westendorf J, Le Clech M, et al.: Plk4-induced centriole biogenesis in human cells. Dev Cell. 2007; 13(2): 190–202. PubMed Abstract | Publisher Full Text\n\nZhao H, Zhu L, Zhu Y, et al.: The Cep63 paralogue Deup1 enables massive de novo centriole biogenesis for vertebrate multiciliogenesis. Nat Cell Biol. 2013; 15(12): 1434–44. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nDzhindzhev NS, Yu QD, Weiskopf K, et al.: Asterless is a scaffold for the onset of centriole assembly. Nature. 2010; 467(7316): 714–8. PubMed Abstract | Publisher Full Text\n\nStrnad P, Leidel S, Vinogradova T, et al.: Regulated HsSAS-6 levels ensure formation of a single procentriole per centriole during the centrosome duplication cycle. Dev Cell. 2007; 13(2): 203–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVulprecht J, David A, Tibelius A, et al.: STIL is required for centriole duplication in human cells. J Cell Sci. 2012; 125(Pt 5): 1353–62. PubMed Abstract | Publisher Full Text\n\nArquint C, Sonnen KF, Stierhof YD, et al.: Cell-cycle-regulated expression of STIL controls centriole number in human cells. J Cell Sci. 2012; 125(Pt 5): 1342–52. PubMed Abstract | Publisher Full Text\n\nHoh RA, Stowe TR, Turk E, et al.: Transcriptional program of ciliated epithelial cells reveals new cilium and centrosome components and links to human disease. PLoS One. 2012; 7(12): e52166. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVladar EK, Stearns T: Molecular characterization of centriole assembly in ciliated epithelial cells. J Cell Biol. 2007; 178(1): 31–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSir JH, Barr AR, Nicholas AK, et al.: A primary microcephaly protein complex forms a ring around parental centrioles. Nat Genet. 2011; 43(11): 1147–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTang TK: Centriole biogenesis in multiciliated cells. Nat Cell Biol. 2013; 15(12): 1400–2. PubMed Abstract | Publisher Full Text\n\nMa L, Quigley I, Omran H, et al.: Multicilin drives centriole biogenesis via E2f proteins. Genes Dev. 2014; 28(13): 1461–71. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBalestra FR, Gönczy P: Multiciliogenesis: multicilin directs transcriptional activation of centriole formation. Curr Biol. 2014; 24(16): R746–9. PubMed Abstract | Publisher Full Text\n\nStubbs JL, Vladar EK, Axelrod JD, et al.: Multicilin promotes centriole assembly and ciliogenesis during multiciliate cell differentiation. Nat Cell Biol. 2012; 14(2): 140–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFunk MC, Bera AN, Menchen T, et al.: Cyclin O (Ccno) functions during deuterosome-mediated centriole amplification of multiciliated cells. EMBO J. 2015; 34(8): 1078–89. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAl Jord A, Lemaître AI, Delgehyr N, et al.: Centriole amplification by mother and daughter centrioles differs in multiciliated cells. Nature. 2014; 516(7529): 104–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKlos Dehring DA, Vladar EK, Werner ME, et al.: Deuterosome-mediated centriole biogenesis. Dev Cell. 2013; 27(1): 103–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKubo A, Sasaki H, Yuba-Kubo A, et al.: Centriolar satellites: molecular characterization, ATP-dependent movement toward centrioles and possible involvement in ciliogenesis. J Cell Biol. 1999; 147(5): 969–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeunier A, Spassky N: Centriole continuity: out with the new, in with the old. Curr Opin Cell Biol. 2016; 38: 60–7. PubMed Abstract | Publisher Full Text\n\nAmemiya CT, Alföldi J, Lee AP, et al.: The African coelacanth genome provides insights into tetrapod evolution. Nature. 2013; 496(7445): 311–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nKramer-Zucker AG, Olale F, Haycraft CJ, et al.: Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer's vesicle is required for normal organogenesis. Development. 2005; 132(8): 1907–21. PubMed Abstract | Publisher Full Text\n\nKemp A: Role of epidermal cilia in development of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi). J Morphol. 1996; 228(2): 203–21. Publisher Full Text\n\nBasti L, Endo M, Segawa S, et al.: Prevalence and intensity of pathologies induced by the toxic dinoflagellate, Heterocapsa circularisquama, in the Mediterranean mussel, Mytilus galloprovincialis. Aquat Toxicol. 2015; 163: 37–50. PubMed Abstract | Publisher Full Text\n\nAzimzadeh J, Wong ML, Downhour DM, et al.: Centrosome loss in the evolution of planarians. Science. 2012; 335(6067): 461–3. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGood MJ, Stommel EW, Stephens RE: Mechanical sensitivity and cell coupling in the ciliated epithelial cells of Mytilus edulis gill. An ultrastructural and developmental analysis. Cell Tissue Res. 1990; 259(1): 51–60. PubMed Abstract | Publisher Full Text\n\nAono K, Fusada A, Fusada Y, et al.: Upside-down gliding of Lymnaea. Biol Bull. 2008; 215(3): 272–9. PubMed Abstract | Publisher Full Text\n\nFerraguti M, Fascio U, Boi S: Mass production of basal bodies in paraspermiogenesis of Tubificinae (Annelida, Oligochaeta). Biol Cell. 2002; 94(2): 109–15. PubMed Abstract | Publisher Full Text\n\nAsh BM, Stephens RE: Ciliogenesis during the sequential formation of molluscan gill filaments. Dev Biol. 1975; 43(2): 340–7. PubMed Abstract | Publisher Full Text\n\nSchermelleh L, Carlton PM, Haase S, et al.: Subdiffraction multicolor imaging of the nuclear periphery with 3D structured illumination microscopy. Science. 2008; 320(5881): 1332–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSchermelleh L, Heintzmann R, Leonhardt H: A guide to super-resolution fluorescence microscopy. J Cell Biol. 2010; 190(2): 165–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVicidomini G, Moneron G, Han KY, et al.: Sharper low-power STED nanoscopy by time gating. Nat Methods. 2011; 8(7): 571–3. PubMed Abstract | Publisher Full Text\n\nLi D, Shao L, Chen BC, et al.: ADVANCED IMAGING. Extended-resolution structured illumination imaging of endocytic and cytoskeletal dynamics. Science. 2015; 349(6251): aab3500. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nXu K, Zhong G, Zhuang X: Actin, spectrin, and associated proteins form a periodic cytoskeletal structure in axons. Science. 2013; 339(6118): 452–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation"
}
|
[
{
"id": "14659",
"date": "28 Jun 2016",
"name": "Steven Brody",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14658",
"date": "28 Jun 2016",
"name": "Giuliano Callaini",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1533
|
https://f1000research.com/articles/5-1532/v1
|
28 Jun 16
|
{
"type": "Review",
"title": "Recent insights: mesenchymal stromal/stem cell therapy for acute respiratory distress syndrome",
"authors": [
"Shahd Horie",
"John G. Laffey",
"Shahd Horie"
],
"abstract": "Acute respiratory distress syndrome (ARDS) causes respiratory failure, which is associated with severe inflammation and lung damage and has a high mortality and for which there is no therapy. Mesenchymal stromal/stem cells (MSCs) are adult multi-progenitor cells that can modulate the immune response and enhance repair of damaged tissue and thus may provide a therapeutic option for ARDS. MSCs demonstrate efficacy in diverse in vivo models of ARDS, decreasing bacterial pneumonia and ischemia-reperfusion-induced injury while enhancing repair following ventilator-induced lung injury. MSCs reduce the pro-inflammatory response to injury while augmenting the host response to bacterial infection. MSCs appear to exert their effects via multiple mechanisms—some are cell interaction dependent whereas others are paracrine dependent resulting from both soluble secreted products and microvesicles/exosomes derived from the cells. Strategies to further enhance the efficacy of MSCs, such as by overexpressing anti-inflammatory or pro-repair molecules, are also being investigated. Encouragingly, early phase clinical trials of MSCs in patients with ARDS are under way, and experience with these cells in trials for other diseases suggests that the cells are well tolerated. Although considerable translational challenges, such as concerns regarding cell manufacture scale-up and issues regarding cell potency and batch variability, must be overcome, MSCs constitute a highly promising potential therapy for ARDS.",
"keywords": [
"ARDS",
"Acute respiratory distress syndrome",
"Mesenchymal stromal/stem cells",
"MSCs"
],
"content": "Introduction\n\nAcute respiratory distress syndrome (ARDS) constitutes a spectrum of increasingly severe acute respiratory failure that may account for 10% of intensive care unit admissions worldwide and continues to have a mortality rate of over 40%1. ARDS can develop in response to multiple predisposing factors, including pneumonia, systemic infection, and major surgery or multiple traumas1. According to the Berlin Definition, the disease can be classified as mild, moderate, or severe, depending on the severity of hypoxemia2. Patients with ARDS have an acute onset of symptoms and signs that include severe dyspnea, chest pain, and cyanosis and bilateral pulmonary infiltrates indicative of edema on chest radiography; these symptoms usually appear within a one-week period from a known clinical insult2.\n\nInflammation is one of the main drivers of ARDS pathogenesis. Damage to the lung endothelial and epithelial layers is caused by inflammation of the lung parenchyma which is mediated by a predominantly neutrophilic influx into the alveolar space3. This leads to the release of pro-inflammatory cytokines, including interleukin-6 (IL-6), IL-1β, and IL-8 and tumor necrosis factor-alpha (TNF-α)3. This in turn leads to the generation of reactive oxygen species that can impair lung barrier function and increase vascular permeability and finally, where ARDS persists, can cause end-stage fibrosis3. Currently, there are no therapies for this condition, and protective mechanical ventilation and fluid-restrictive strategies are the only supportive treatments that can improve outcome. Mesenchymal stromal/stem cells (MSCs) have attracted considerable interest as a potential therapy for ARDS. In this review, we will discuss early key studies that first demonstrated the potential of MSCs for ARDS and subsequently focus on recent findings and advances published within the last 2 years.\n\n\nRationale for mesenchymal stromal/stem cell therapy for acute respiratory distress syndrome\n\nMSCs are multipotent adult progenitor cells that can be isolated from numerous sources, including bone marrow, umbilical cord, and adipose tissue, and can differentiate into mesenchymal lineage cells such as fat or cartilage tissue4. The interest in MSCs as a possible therapy for ARDS stems largely from their potential to modulate the host immune response to injury and infection and to promote repair following tissue injury. A number of seminal studies showed the therapeutic potential of MSCs for ARDS. Gupta et al. demonstrated that MSCs significantly improve survival and reduced lung injury in mice following Escherichia coli endotoxin lung injury5. Importantly, MSCs modulated the immune response by inhibiting TNF-α release and by enhancing anti-inflammatory IL-10 secretion5. Mei et al. demonstrated that MSCs genetically modified to overexpress angiopoietin 1, an important protein involved in vascular growth and angiogenesis, were more effective than naïve MSC treatment in a mouse model of lipopolysaccharide (LPS)-induced lung injury6. Chang et al. demonstrated that human umbilical cord blood-derived MSCs attenuated hyperoxia-induced lung injury in the neonatal rat, a key demonstration of the efficacy of MSCs from other, potentially more plentiful, tissue sources7. Nemeth et al. demonstrated the therapeutic potential of MSCs in a murine model of systemic sepsis, and efficacy was mediated via reprogramming of macrophage function8. Of translational significance, Lee et al. demonstrated that human MSCs and MSC conditioned medium (MSC-CM) decreased endotoxin-induced injury to the human ex vivo perfused lung via mechanisms involving improved alveolar fluid clearance9. Krasnodembskaya et al. highlighted the role of MSC-secreted antimicrobial peptides in augmenting the host response to bacterial infection10. Curley et al. demonstrated the potential for MSC therapy to repair the lung following ventilation-induced acute lung injury (ALI)11.\n\nThe immune-modulating effect of MSCs is increasingly well understood. MSCs promote T regulatory cell expansion, which causes the suppression of the proliferation of effector T cells and dampens the immune response12. Furthermore, MSCs can modify T cells, dendritic cells, and natural killer cells and decrease their release of pro-inflammatory cytokines or increase their release of anti-inflammatory molecules12. MSCs can also cause the release of soluble potent anti-inflammatory mediators such as indoleamine 2,3-dioxygenase (IDO), prostaglandin E2 (PGE2), and IL-1012,13. In terms of their reparative function, MSCs secrete growth factors such as keratinocyte growth factor (KGF) and vascular endothelial growth factor (VEGF)14,15. Importantly for an acute inflammatory disorder such as ARDS, MSCs do not elicit a significant immune response in the host and thus can be used for allogeneic transplantation12.\n\n\nRecent insights regarding mesenchymal stromal/stem cells in preclinical acute respiratory distress syndrome models\n\nMSCs have previously shown promising efficacy in numerous preclinical ARDS models, including pneumonia, ventilator-induced lung injury (VILI), and sepsis. A recent study showed that mouse MSCs significantly improved survival, reduced histologic lung injury, and reduced pulmonary edema in a mouse model of E. coli-induced ALI via a mechanism involving reduced oxidant stress (Table 1)16. Devaney et al. also recently demonstrated that human bone marrow (hBM)-derived MSCs reduce injury severity and increase survival in a rat model of E. coli pneumonia (Table 1)17. hBM-MSCs reduced the lung bacterial load following intra-tracheal E. coli administration via a mechanism involving enhanced antimicrobial peptide secretion and increased macrophage phagocytosis17. hBM-MSCs decreased alveolar concentrations of IL-6 while increasing concentrations of the anti-inflammatory cytokine IL-10 and increasing KGF17. Other findings of translational significance included delineation of the human MSC dose response curve and the demonstration that cryopreserved cells retain efficacy, but the secretome alone was less effective in the setting of E. coli pneumonia17. In a key study, Asmussen et al. showed that clinical-grade human MSCs significantly enhanced oxygenation while reducing pulmonary edema without any adverse effects in a large animal (sheep) model of combined smoke inhalation and Pseudomonas aeruginosa pneumonia (Table 1)18. The translational importance of this study is underlined by the fact that these cells are now in early phase clinical trials for ARDS.\n\nAbbreviations: ACE2, angiotensin-converting enzyme 2; ALI, acute lung injury; CINC-1, cytokine-induced neutrophil chemoattractant 1; hBM, human bone marrow; IL, interleukin; IV, intravenously; MSC, mesenchymal stromal/stem cell; TNF-α, tumor necrosis factor-alpha.\n\nPrevious studies observed MSCs to enhance lung repair following VILI. Chimenti et al. demonstrated the potential for MSC pretreatment to prevent the development of VILI following high-volume ventilation. MSC pretreatment reduced the lung water content and improved the lung histology score19. Bronchoalveolar lavage (BAL) levels of protein, neutrophils, macrophage inflammatory protein-2, and IL-1β were also considerably reduced7. Curley et al. observed that intra-tracheal and intravenously delivered rat MSCs and their secretome comparably enhanced restoration of lung oxygenation and compliance and restored lung structure after high-pressure VILI20. Inflammation was also significantly attenuated by MSC delivery as indicated by lower concentrations of IL-6 and TNF-α in the BAL20. More recently, Hayes et al. showed that MSCs retained their therapeutic efficacy following high-pressure VILI at doses as low as 2 million cells per kilogram (Table 1)21. They also observed that efficacy with MSC treatment was maintained with delayed delivery (that is, 6 hours after injury)21.\n\nMSCs enhance alveolar fluid clearance, an important process in the recovery following ARDS. Bacteria-induced inflammation was also significantly inhibited and this appeared to be due to the enhanced phagocytic ability of macrophages and to antibacterial effects of secreted KGF in an ex vivo perfused human lung pneumonia model22. Of importance in transplant medicine, MSCs also restored alveolar fluid clearance in another study which used ex vivo lungs that were ischemic and rejected for transplantation (Table 1)23.\n\n\nStrategies to enhance mesenchymal stromal/stem cell efficacy\n\nA key early study demonstrated the potential to enhance MSC efficacy by using gene overexpression strategies, specifically overexpression of Ang-16. More recently, MSCs overexpressing angiotensin-converting enzyme 2 (ACE2) showed enhanced endothelial repair and enhanced reduction in the expression of inflammatory molecules (including TNF-α and IL-6) following in vitro LPS injury when compared with naïve MSCs24 (Table 1). ACE2 plays a major role in negating the pro-inflammatory and pro-apoptotic effects of angiotensin II on the endothelium24. In vivo, Min et al. showed that MSCs overexpressing ACE2 caused an enhanced reduction in inflammation and reduced lung edema, collagen deposition, and fibrosis after bleomycin-induced lung injury in mice25. Furthermore, these cells decreased oxidative stress damage to a greater degree than naïve MSCs25. Finally, human MSCs overexpressing soluble IL-1 receptor-like-1, which plays an immune-modulatory role in the lung, decreased inflammation, BAL protein and neutrophil content to a greater extent than naïve MSCs in a mouse model of LPS-induced ALI26.\n\n\nRecent insights regarding the mesenchymal stromal/stem cell secretome\n\nMSCs exert their therapeutic efficacy in part via paracrine mechanisms (Table 2). MSC-CM—termed the ‘MSC secretome’—attenuates injury following endotoxin-induced ALI in the mouse27, restoring lung structure and decreasing alveolar concentrations of IL-6 and macrophage inflammatory protein-2. These effects were partly mediated via the attenuation of the expression of nuclear factor-kappa B (NF-κB) p65 and phospho-NF-κB p65 in the mouse lung27. MSC-CM also induced neutrophil apoptosis as indicated by a reduction of the expression of Bcl-xl and Mcl-1 on said neutrophils27. In a similar model, MSC-CM enhanced the activation of the M2 macrophage phenotype, thus promoting an anti-inflammatory and pro-healing environment28. Interestingly, these effects were correlated with increased levels of insulin-like growth factor 1 (IGF-1)18.\n\nAbbreviations: ALI, acute lung injury; MSC, mesenchymal stromal/stem cell; NF-κB, nuclear factor-kappa B; KGF, keratinocyte growth factor; BAL, bronchoalveolar lavage.\n\nThe potential for the MSC secretome to enhance lung repair following VILI has also been demonstrated. Intra-tracheal delivery of the MSC secretome products enhanced lung oxygenation and compliance and reduced the lung histology score and pulmonary edema at 48 hours after high-pressure VILI20. Another study examined the effects of MSCs and their secretome in the earlier phases of the repair process following VILI29. The MSCs were effective and significantly improved blood oxygenation and respiratory compliance while reducing lung edema, BAL neutrophil counts, and IL-6 concentrations29. In contrast, the MSC secretome failed to exert therapeutic effects at this earlier stage in the recovery process, suggesting that not all of the effects of the MSC are recapitulated by the secretome alone29.\n\n\nRecent insights regarding mesenchymal stromal/stem cell-derived microvesicles and exosomes\n\nA relatively new discovery is that some of the effects of MSCs appear to be mediated via the release of particles, specifically microvesicles and exosomes that can transfer genetic material, cellular organelles such as mitochondria, and biologically active molecules between cells (Table 2). One study observed that microvesicles from hBM-MSCs significantly decreased lung edema and BAL protein and neutrophil counts in an LPS lung injury model and this was facilitated partly by the increased expression of KGF30. Islam et al. demonstrated that MSCs delivered to LPS-injured mice transferred mitochondria-containing microvesicles through gap junctions to the mouse alveolar epithelia and that the therapeutic effect was abrogated with the use of MSCs with non-functional mitochondria or gap junction-incompetent MSCs31. Another study observed that human MSC-derived microvesicles enhanced survival in a mouse model of E. coli pneumonia32. The microvesicles were comparable to MSC cell treatment in their ability to reduce bacterial load and BAL protein and cytokine concentrations, and their effects were mediated in part by KGF secretion32. Furthermore, the phagocytic capability of monocytes in vitro was also increased by the microvesicles32. Jackson et al. also observed that MSCs enhanced macrophage phagocytosis in vitro and produced anti-bacterial effects in an E. coli pneumonia mouse model in part via microvesicle transfer of mitochondria33.\n\n\nClinical studies\n\nBased on their therapeutic promise in preclinical studies, MSCs have entered early phase clinical testing in patients with ARDS (Table 3). The recently reported phase 1b START trial (safety proof-of-concept trial) enrolled nine patients to receive 1, 5, or 10 million cells per kilogram of a single intravenous dose of allogeneic hBM MSCs using a three-by-three dose escalation design and showed no adverse effects at any of the doses used34. These investigators are now enrolling patients in a phase 2 study using the dose of 10 million cells per kilogram (NCT01775774). In another phase 1 trial, adipose-derived allogeneic MSCs were administered to 12 patients in an intravenous dose of 1 million cells per kilogram or placebo in a 1:1 ratio and again the cells were well tolerated in the patients35. This study also measured IL-6, IL-8, and surfactant protein D levels and found that surfactant protein D concentrations were lower after MSC delivery at day 5 but that MSC administration had no effects on IL-6 or IL-8 levels35. The study concluded that higher doses may be needed at the phase 2 stage35. Importantly, given the relationship between ARDS and sepsis, MSCs are also in early phase clinical studies in patients with septic shock in Canada (Cellular Immunotherapy in Septic Shock study, NCT02421484).\n\n\nMesenchymal stromal/stem cells for acute respiratory distress syndrome: challenges and next steps\n\nAlthough MSCs demonstrate considerable promise and early phase clinical trials are encouraging, several challenges still impede the clinical translation of MSCs for patients with ARDS. Cell batch variability between donors remains a significant hurdle that raises efficacy and possible safety concerns. Robust assays of MSC batch potency for ARDS are lacking. However, one recent study observed that cell motility can predict the differentiation potential of MSCs, which may provide a bioassay for characterizing MSC efficacy for certain disease conditions36. Another study also recently outlined that MSCs from young mouse donors were more effective than those from aging donors in reducing pulmonary fibrosis37. Of relevance to ARDS, MSCs from aging mice were less effective in reducing endotoxemia-induced lung injury38. Importantly, the efficacy of these cells in an in vitro cell motility assay paralleled their effectiveness in subsequent in vivo studies, suggesting that this may be a useful assay of potency38.\n\nFurthermore, the generation of large quantities of clinical-grade MSCs for dose delivery to humans is both time consuming and costly. Doses of up to 1 billion MSCs per patient are currently being studied in clinical trials. Should MSCs prove an effective therapy at these doses, alternative, more easily accessible and abundant sources, such as adipose tissue or the umbilical cord, are likely to prove more feasible than bone marrow, which is the current source of cells used in most clinical studies. The demonstration that MSCs retain their efficacy when cryopreserved is important in making bulk production and storage of MSCs for clinical use feasible17 and will facilitate MSC translation to the clinic.\n\nIn conclusion, MSCs may provide the immunomodulatory and regenerative tool needed to target ARDS pathogenesis. Preclinical studies have highlighted their therapeutic potential and mechanisms of action. These preclinical studies may also be of relevance to other clinical situations, such as smoke inhalation injury, lung transplantation, and chemotherapy-induced lung injury. Although there are considerable grounds for optimism regarding the therapeutic potential of MSCs for ARDS, hurdles to translation still exist, and only time will tell whether MSCs realize their therapeutic potential in later-phase clinical trials.\n\n\nAbbreviations\n\nACE2, angiotensin-converting enzyme 2; ALI, acute lung injury; ARDS, acute respiratory distress syndrome; BAL, bronchoalveolar lavage; hBM, human bone marrow; IL, interleukin; KGF, keratinocyte growth factor; LPS, lipopolysaccharide; MSC, mesenchymal stromal/stem cell; MSC-CM, mesenchymal stromal/stem cell conditioned medium; NF-κB, nuclear factor-kappa B; TNF-α, tumor necrosis factor-alpha; VILI, ventilator-induced lung injury.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nShahd Horie is supported by funding from the Health Research Board, Ireland. John G. Laffey is funded by the Canadian Institute of Health Research and Physicians Services Incorporated, Ontario, and holds a Merit Award from the University of Toronto, Department of Anesthesia.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBellani G, Laffey JG, Pham T, et al.: Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. JAMA. 2016; 315(8): 788–800. PubMed Abstract | Publisher Full Text\n\nARDS Definition Task Force, Ranieri VM, Rubenfeld GD, et al.: Acute respiratory distress syndrome: the Berlin Definition. JAMA. 2012; 307(23): 2526–33. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFanelli V, Vlachou A, Ghannadian S, et al.: Acute respiratory distress syndrome: new definition, current and future therapeutic options. J Thorac Dis. 2013; 5(3): 326–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing DC, Shyu WC, Lin SZ: Mesenchymal stem cells. Cell Transplant. 2011; 20(1): 5–14. PubMed Abstract | Publisher Full Text\n\nGupta N, Su X, Popov B, et al.: Intrapulmonary delivery of bone marrow-derived mesenchymal stem cells improves survival and attenuates endotoxin-induced acute lung injury in mice. J Immunol. 2007; 179(3): 1855–63. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMei SH, McCarter SD, Deng Y, et al.: Prevention of LPS-induced acute lung injury in mice by mesenchymal stem cells overexpressing angiopoietin 1. PLoS Med. 2007; 4(9): e269. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nChang YS, Oh W, Choi SJ, et al.: Human umbilical cord blood-derived mesenchymal stem cells attenuate hyperoxia-induced lung injury in neonatal rats. Cell Transplant. 2009; 18(8): 869–86. PubMed Abstract | Publisher Full Text\n\nNemeth K, Leelahavanichkul A, Yuen PS, et al.: Bone marrow stromal cells attenuate sepsis via prostaglandin E2-dependent reprogramming of host macrophages to increase their interleukin-10 production. Nat Med. 2009; 15(1): 42–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nLee JW, Fang X, Gupta N, et al.: Allogeneic human mesenchymal stem cells for treatment of E. coli endotoxin-induced acute lung injury in the ex vivo perfused human lung. Proc Natl Acad Sci U S A. 2009; 106(38): 16357–62. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nKrasnodembskaya A, Song Y, Fang X, et al.: Antibacterial effect of human mesenchymal stem cells is mediated in part from secretion of the antimicrobial peptide LL-37. Stem Cells. 2010; 28(12): 2229–38. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCurley GF, Hayes M, Ansari B, et al.: Mesenchymal stem cells enhance recovery and repair following ventilator-induced lung injury in the rat. Thorax. 2012; 67(6): 496–501. PubMed Abstract | Publisher Full Text\n\nAggarwal S, Pittenger MF: Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood. 2005; 105(4): 1815–22. PubMed Abstract | Publisher Full Text\n\nBeyth S, Borovsky Z, Mevorach D, et al.: Human mesenchymal stem cells alter antigen-presenting cell maturation and induce T-cell unresponsiveness. Blood. 2005; 105(5): 2214–9. PubMed Abstract | Publisher Full Text\n\nAguilar S, Scotton CJ, McNulty K, et al.: Bone marrow stem cells expressing keratinocyte growth factor via an inducible lentivirus protects against bleomycin-induced pulmonary fibrosis. PLoS One. 2009; 4(11): e8013. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiacca M, Zacchigna S: VEGF gene therapy: therapeutic angiogenesis in the clinic and beyond. Gene Ther. 2012; 19(6): 622–9. PubMed Abstract | Publisher Full Text\n\nShalaby SM, El-Shal AS, Abd-Allah SH, et al.: Mesenchymal stromal cell injection protects against oxidative stress in Escherichia coli-induced acute lung injury in mice. Cytotherapy. 2014; 16(6): 764–75. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nDevaney J, Horie S, Masterson C, et al.: Human mesenchymal stromal cells decrease the severity of acute lung injury induced by E. coli in the rat. Thorax. 2015; 70(7): 625–35. PubMed Abstract | Publisher Full Text\n\nAsmussen S, Ito H, Traber DL, et al.: Human mesenchymal stem cells reduce the severity of acute lung injury in a sheep model of bacterial pneumonia. Thorax. 2014; 69(9): 819–25. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nChimenti L, Luque T, Bonsignore MR, et al.: Pre-treatment with mesenchymal stem cells reduces ventilator-induced lung injury. Eur Respir J. 2012; 40(4): 939–48. PubMed Abstract | Publisher Full Text\n\nCurley GF, Ansari B, Hayes M, et al.: Effects of intratracheal mesenchymal stromal cell therapy during recovery and resolution after ventilator-induced lung injury. Anesthesiology. 2013; 118(4): 924–32. PubMed Abstract | Publisher Full Text\n\nHayes M, Masterson C, Devaney J, et al.: Therapeutic efficacy of human mesenchymal stromal cells in the repair of established ventilator-induced lung injury in the rat. Anesthesiology. 2015; 122(2): 363–73. PubMed Abstract | Publisher Full Text\n\nLee JW, Krasnodembskaya A, McKenna DH, et al.: Therapeutic effects of human mesenchymal stem cells in ex vivo human lungs injured with live bacteria. Am J Respir Crit Care Med. 2013; 187(7): 751–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcAuley DF, Curley GF, Hamid UI, et al.: Clinical grade allogeneic human mesenchymal stem cells restore alveolar fluid clearance in human lungs rejected for transplantation. Am J Physiol Lung Cell Mol Physiol. 2014; 306(9): L809–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHe HL, Liu L, Chen QH, et al.: MSCs modified with ACE2 restore endothelial function following LPS challenge by inhibiting the activation of RAS. J Cell Physiol. 2015; 230(3): 691–701. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMin F, Gao F, Li Q, et al.: Therapeutic effect of human umbilical cord mesenchymal stem cells modified by angiotensin-converting enzyme 2 gene on bleomycin-induced lung fibrosis injury. Mol Med Rep. 2015; 11(4): 2387–96. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMartinez-Gonzalez I, Roca O, Masclans JR, et al.: Human mesenchymal stem cells overexpressing the IL-33 antagonist soluble IL-1 receptor-like-1 attenuate endotoxin-induced acute lung injury. Am J Respir Cell Mol Biol. 2013; 49(4): 552–62. PubMed Abstract | Publisher Full Text\n\nSu VY, Yang KY: Mesenchymal stem cell-conditioned medium induces neutrophils apoptosis via inhibition of NF-kB pathway and increases endogenous pulmonary stem cells in endotoxin-induced acute lung injury. European Respiratory Journal. 2015, 46(suppl 59): OA3520. Publisher Full Text\n\nIonescu L, Byrne RN, van Haaften T, et al.: Stem cell conditioned medium improves acute lung injury in mice: in vivo evidence for stem cell paracrine action. Am J Physiol Lung Cell Mol Physiol. 2012; 303(11): L967–77. PubMed Abstract | Free Full Text | F1000 Recommendation\n\nHayes M, Curley GF, Masterson C, et al.: Mesenchymal stromal cells are more effective than the MSC secretome in diminishing injury and enhancing recovery following ventilator-induced lung injury. Intensive Care Med Exp. 2015; 3(1): 29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhu YG, Feng XM, Abbott J, et al.: Human mesenchymal stem cell microvesicles for treatment of Escherichia coli endotoxin-induced acute lung injury in mice. Stem Cells. 2014; 32(1): 116–25. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nIslam MN, Das SR, Emin MT, et al.: Mitochondrial transfer from bone-marrow-derived stromal cells to pulmonary alveoli protects against acute lung injury. Nat Med. 2012; 18(5): 759–65. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMonsel A, Zhu YG, Gennai S, et al.: Therapeutic Effects of Human Mesenchymal Stem Cell-derived Microvesicles in Severe Pneumonia in Mice. Am J Respir Crit Care Med. 2015; 192(3): 324–36. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nJackson MV, Morrison TJ, O’Kane CM, et al.: T3 Mitochondrial transfer is an important mechanism by which Mesenchymal Stromal Cells (MSC) facilitate macrophage phagocytosis in the in vitro and in vivo models of Acute Respiratory Distress Syndrome (ARDS). Thorax. 2015; 70(Suppl 3): A1.3–A2. Publisher Full Text | F1000 Recommendation\n\nWilson JG, Liu KD, Zhuo H, et al.: Mesenchymal stem (stromal) cells for treatment of ARDS: a phase 1 clinical trial. Lancet Respir Med. 2015; 3(1): 24–32. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nZheng G, Huang L, Tong H, et al.: Treatment of acute respiratory distress syndrome with allogeneic adipose-derived mesenchymal stem cells: a randomized, placebo-controlled pilot study. Respir Res. 2014; 15(1): 39. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBertolo A, Gemperli A, Gruber M, et al.: In vitro cell motility as a potential mesenchymal stem cell marker for multipotency. Stem Cells Transl Med. 2015; 4(1): 84–90. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nTashiro J, Elliot SJ, Gerth DJ, et al.: Therapeutic benefits of young, but not old, adipose-derived mesenchymal stem cells in a chronic mouse model of bleomycin-induced pulmonary fibrosis. Transl Res. 2015; 166(6): 554–67. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBustos ML, Huleihel L, Kapetanaki MG, et al.: Aging mesenchymal stem cells fail to protect because of impaired migration and antiinflammatory response. Am J Respir Crit Care Med. 2014; 189(7): 787–98. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation"
}
|
[
{
"id": "14660",
"date": "28 Jun 2016",
"name": "Duncan Stewart",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14661",
"date": "28 Jun 2016",
"name": "Michael O'Connor",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1532
|
https://f1000research.com/articles/5-950/v1
|
20 May 16
|
{
"type": "Software Tool Article",
"title": "Infrastructure for genomic interactions: Bioconductor classes for Hi-C, ChIA-PET and related experiments",
"authors": [
"Aaron T. L. Lun",
"Malcolm Perry",
"Elizabeth Ing-Simmons",
"Malcolm Perry",
"Elizabeth Ing-Simmons"
],
"abstract": "The study of genomic interactions has been greatly facilitated by techniques such as chromatin conformation capture with high-throughput sequencing (Hi-C). These genome-wide experiments generate large amounts of data that require careful analysis to obtain useful biological conclusions. However, development of the appropriate software tools is hindered by the lack of basic infrastructure to represent and manipulate genomic interaction data. Here, we present the InteractionSet package that provides classes to represent genomic interactions and store their associated experimental data, along with the methods required for low-level manipulation and processing of those classes. The InteractionSet package exploits existing infrastructure in the open-source Bioconductor project, while in turn being used by Bioconductor packages designed for higher-level analyses. For new packages, use of the functionality in InteractionSet will simplify development, allow access to more features and improve interoperability between packages.",
"keywords": [
"Hi-C",
"ChIA-PET",
"infrastructure",
"data representation",
"genomic interactions"
],
"content": "Introduction\n\nTechniques such as chromatin conformation capture with high-throughput sequencing (Hi-C)1 and chromatin interaction analysis with paired-end tags (ChIA-PET)2 are increasingly being used to study the three-dimensional structure and organisation of the genome. Briefly, genomic DNA is fragmented and subjected to a ligation step during which DNA from interacting loci are ligated together. High-throughput paired-end sequencing of the ligation products will identify pairs of interacting genomic regions. The strength of each interaction can also be quantified from the number of read pairs connecting the two interacting regions. This information can be used to derive biological insights into the role of long-range interactions in transcriptional regulation as well as the general organization of the genome inside the nucleus.\n\nThe analysis of Hi-C and ChIA-PET data is not a trivial task, and many software packages have been developed to facilitate this process. Several of these packages like diffHic3 and GenomicInteractions4 are part of the open-source Bioconductor project, which aims to provide accessible tools for analyzing high-throughput genomic data with the R programming language. One of the strengths of the Bioconductor project is the quality and quantity of shared infrastructure available to developers5. Pre-defined S4 classes such as GenomicRanges and SummarizedExperiment can be used to represent various types of genomic data and information, easing the maintenance burden for developers while also improving interoperability between packages for users. However, this kind of common infrastructure does not yet exist for the genomic interaction field. Instead, each package contains its own custom classes, which increases code redundancy and development load while reducing interoperability.\n\nHere, we describe the InteractionSet package that provides base S4 classes for representing and manipulating genomic interaction data. It contains the GInteractions class, to represent pairwise interactions; the InteractionSet class, to store the associated experimental data; and the ContactMatrix class, to represent interactions in a matrix format. This facilitates code reuse across Bioconductor packages involved in analyzing data from Hi-C, ChIA-PET and similar experiments.\n\n\nOverview of available classes\n\nEach object of the GInteractions class is designed to represent interactions between pairs of “anchor” regions in the genome (Figure 1A). It does so by storing pairs of anchor indices that point towards a reference set of genomic coordinates (specified as a GenomicRanges object). Each anchor index refers to a specific reference region, such that a pair of such indices represents a pairwise interaction between the corresponding regions. This design reduces memory usage as the reference coordinates need only be stored once, even if each region is involved in multiple interactions. Computational work is also reduced as calculations can be quickly applied across the small set of reference regions, and the results can be retrieved for each interaction based on the anchor indices. In addition, the GInteractions class inherits from the Vector class in Bioconductor’s S4Vectors package. This allows storage of metadata for each interaction (e.g., intensities, p-values) and for the entire object (e.g., experiment description).\n\nThe InteractionSet class is designed to store experimental data for each feature (Figure 1B). It inherits from the SummarizedExperiment base class, where each object of the class stores any number of matrices of the same dimensions. Each row of each matrix corresponds to a pairwise genomic interaction (represented by a GInteractions object that is also stored within each InteractionSet object), while each column corresponds to an experimental sample. Each entry of the matrix then represents the observation for the corresponding interaction in the corresponding sample. Different matrices can be used to store different types of data, e.g., read counts, normalized intensities. The InteractionSet class also inherits a number of fields to store metadata for each interaction, for each sample, and for the entire object.\n\nThe ContactMatrix class is designed to represent pairwise interactions in a matrix format (Figure 1C). Each row and column of the matrix represents a genomic region, such that each cell of the matrix represents an interaction between the corresponding row/column regions. Experimental data for that interaction can be stored in the associated cell. This provides a direct representation of the “interaction space”, i.e., the two-dimensional space in which (x, y) represents an interaction between x and y. Like the GInteractions class, the genomic coordinates are not stored directly – rather, the rows/columns have indices that point towards a reference set of coordinates, which reduces memory usage and computational work. The matrix representation itself uses classes in the Matrix package to provide support for both dense and sparse matrices. The latter may be more memory-efficient, particularly for sparse areas of the interaction space. The ContactMatrix class is compatible with existing matrix-based classes such as those in the HiTC package6.\n\n\nOverview of available methods\n\nThe InteractionSet package provides a variety of methods for manipulating objects of each class. In addition to slot accessors and modifiers, methods are available to convert objects to different classes in the same package (e.g., GInteractions to ContactMatrix) or to base Bioconductor classes (e.g., GInteractions to GRangesList). The distance between anchor regions on the linear genome can be computed for each pairwise interaction, to use in fitting a distance-dependent trend1 for diagnostics or normalization. The minimum bounding box in the interaction space can also be defined for a group of interactions (Figure 2A) to summarize the location of that group.\n\nRelevant slots of each class (i.e., data values stored in each object of the class) are labelled with a preceding “@”. (A) The GInteractions class represents pairwise interactions between genomic regions by storing pairs of anchor indices that refer to coordinates in a GenomicRanges object. (B) The InteractionSet class stores experimental data in an “assays” matrix where each row is an interaction and each column is a sample. Here, counts represent the number of read pairs mapped between each pair of interacting regions in each sample. (C) The ContactMatrix class represents the interaction space as a matrix, where each cell represents an interaction between the corresponding row/column regions.\n\nThe InteractionSet package supports one- or two-dimensional overlaps for its objects (Figure 2B). A one-dimensional overlap is considered to be present between an interaction and a genomic interval if either anchor region of the interaction overlaps the interval. This can be used to identify interactions overlapping pre-defined regions of interest. A two-dimensional overlap is considered to be present between an interaction and two genomic intervals if one anchor region overlaps one interval and the other anchor region overlaps the other interval. This can be used to identify interactions linking two specific regions of interest, e.g., a gene and its enhancer. The same framework can be used to define two-dimensional overlaps between two interactions, based on whether the corresponding anchor regions overlap – this can be used to relate similar interactions in different GInteractions objects or across different experiments. More generally, interactions can be identified that link any two regions in a set of regions of interest. For example, given a set of genes, interactions between two genes can be identified; or given a set of genes and another set of enhancers, interactions linking any gene to any enhancer can be found.\n\nHi-C data in an InteractionSet object can also be converted into a 4C-like format (Figure 2C). Firstly, a bait region is defined as some region of interest, e.g., a target gene or enhancer. All interactions in the InteractionSet object that have one-dimensional overlaps with the bait are identified. For each overlapping interaction, the anchor region that does not overlap with the bait is extracted and – along with the data associated with that interaction – used to construct a RangedSummarizedExperiment object. This process yields data for intervals on the linear genome, which is similar to the output of 4C experiments7 that measure the intensity of interactions between the bait and all other regions. The “linearized” format may be preferable when a specific region can be defined as the bait, as intervals on the linear genome are easier to interpret than interactions in two-dimensional space.\n\n(A) Minimum bounding boxes can be identified for groups of interactions using the boundingBox method. Here, u′, v′ and w′ belong in one group while x′, y′ and z′ belong in another. (B) One- or two-dimensional overlaps can be identified between interactions and one or two genomic intervals, respectively, using the findOverlaps method. Here, x′ and y′ have one-dimensional overlaps with the gene and enhancer, respectively, while z′ has a two-dimensional overlap with the gene and the enhancer. (C) An InteractionSet object contains data – in this case, read pair count data – for interactions in the two-dimensional interaction space. Given a bait region, a “cross-section” of the space can be extracted and converted into a RangedSummarizedExperiment object using the linearize method. This object holds count data for intervals on the linear genome (blue lines) where the count for each interval describes the strength of the interaction between that interval and the bait. This format effectively mimics that of 4C data.\n\n\nImplementation and operation details\n\nAll classes and methods in the InteractionSet package are implemented using the S4 object-orientated framework in R. Classes are exported to allow package developers to derive custom classes for their specific needs. Pre-existing Bioconductor classes and generics are used to provide a consistent interface for users. After loading the InteractionSet package into an R session, instances of each class can be constructed from existing data structures, either directly (e.g., GInteractions objects from GRanges via the GInteractions constructor, or from Pairs via the makeGInteractionsfromGRanges function; ContactMatrix objects from GRanges and Matrix via the ContactMatrix constructor) or in a hierarchical manner (e.g., InteractionSet objects from matrices and a GInteractions object via the InteractionSet constructor). The methods described above can then be applied to each instance of the class. While the InteractionSet package does not have functions to load data from file, it can be combined with the import function in the rtracklayer package8 to construct class instances after importing data from a range of formats including BED and BEDPE. A similar strategy can be used to export data to file.\n\n\nConclusions\n\nThe availability of common infrastructure is highly beneficial to software development by reducing redundancy and improving reliability, as more developers can check the same code; improving interoperability, as different packages use the same classes; and increasing the accessibility of useful features, which exist in a single package rather than being sequestered away in a variety of different packages. Here, we present the InteractionSet package that implements a number of classes and methods for representing, storing and manipulating genomic interaction data from Hi-C, ChIA-PET and related experiments. The package is fully integrated into the Bioconductor ecosystem, depending on a number of base packages to implement its classes (e.g., S4Vectors, GenomicRanges, SummarizedExperiment) while in turn being depended on by packages for higher-level analyses (e.g., diffHic, GenomicInteractions). Indeed, for any new packages, use of the features in InteractionSet will simplify development and improve interoperability with existing packages in the Bioconductor project. The InteractionSet package itself can be obtained for R version 3.3.0 at http://bioconductor.org/packages/InteractionSet.\n\n\nSoftware availability\n\nSoftware and latest source code available from: http://bioconductor.org/packages/InteractionSet\n\nArchived source code as at time of publication: http://dx.doi.org/10.5281/zenodo.512049\n\nLicense: GNU General Public License version 3.0",
"appendix": "Author contributions\n\n\n\nATLL proposed and developed the InteractionSet package, with significant contributions from MP and EI-S. All authors wrote and approved the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nATLL was supported by core funding from Cancer Research UK (award no. A17197). MP and EI-S were supported by a Medical Research Council PhD studentships.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Annika Gable, Aleksandra Pekowska, Bernd Klaus, Michael Lawrence and Hervé Pagès for coding and feature suggestions. We also thank John Marioni and Boris Lenhard for comments on the manuscript.\n\n\nReferences\n\nLieberman-Aiden E, van Berkum NL, Williams L, et al.: Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science. 2009; 326(5950): 289–293. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFullwood MJ, Liu MH, Pan YF, et al.: An oestrogen-receptor-alpha-bound human chromatin interactome. Nature. 2009; 462(7269): 58–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Smyth GK: diffHic: a Bioconductor package to detect differential genomic interactions in Hi-C data. BMC Bioinformatics. 2015; 16(1): 258. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarmston N, Ing-Simmons E, Perry M, et al.: GenomicInteractions: An R/Bioconductor package for manipulating and investigating chromatin interaction data. BMC Genomics. 2015; 16(1): 963. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nServant N, Lajoie BR, Nora EP, et al.: HiTC: exploration of high-throughput ‘C’ experiments. Bioinformatics. 2012; 28(21): 2843–2844. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimonis M, Klous P, Splinter E, et al.: Nuclear organization of active and inactive chromatin domains uncovered by chromosome conformation capture-on-chip (4C). Nat Genet. 2006; 38(11): 1348–1354. PubMed Abstract | Publisher Full Text\n\nLawrence M, Gentleman R, Carey V: rtracklayer: an R package for interfacing with genome browsers. Bioinformatics. 2009; 25(14): 1841–1842. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun A, Perry M, Ing-Simmons E, et al.: Base Classes for Storing Genomic Interaction Data. Zenodo. 2016. Publisher Full Text"
}
|
[
{
"id": "13924",
"date": "25 May 2016",
"name": "Douglas Phanstiel",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe an R package that allows users to store, organize, manipulate, and intersect pairwise chromosomal interactions. The package is written for use with HiC and ChIA-PET style data sets but could be used for a variety of pairwise interactions. While there exist a plethora of tools for the storage, manipulation, and analysis of single genomic elements, equivalent tools adapted for pairwise genomic interactions are limited and still greatly needed.\nThe paper is well written, simple, and accurately describes the package itself. I have downloaded and tested the package and both download and usage went smoothly. It behaves similarly to some of the packages that it builds off of such as GenomicRanges. And those familiar with GenomicRanges will be at home when using InteractionSet. In addition to simply storing and organizing pairwise data, InteractionSet includes a lot of handy features that will be of great use to the community. These include simple functions that organize pairwise genomic interactions such as swapAnchors that assures that the first of the two paired regions is always on the lower numbered chromosome or upstream (with regard to Watson strand) and more complex functions such as findOverlaps that allows users to overlap sets of pairwise interactions in a variety of ways.\n\nThis package is very useful and powerful and provides a valuable resource to software developers and advanced users. The GenomicRanges-style organization of the data, that InteractionSet adopts, is often too complicated for casual R users to learn. In many cases simply reading files from BED, BEDPE, or sam format into data frames is easier and faster for simple tasks. However, developers will prefer this more standardized format for improved stability of their packages. And advanced users may prefer the standardized yet flexible approach to data organization and the powerful built in tools.\nImportantly, the package is accompanied by a detailed and clear online tutorial which clearly demonstrates how to use the classes and functions. In summary, this paper is succinct and clearly written and accurately describes an R package that will be of great use to the scientific community.",
"responses": [
{
"c_id": "2007",
"date": "28 Jun 2016",
"name": "Aaron Lun",
"role": "Author Response",
"response": "Thanks for your comments Douglas. We agree that the Bioconductor ecosystem of data classes can be somewhat daunting for new users. Nonetheless, we believe that the use of standard Bioconductor tools is the safest strategy for the majority of users (and obviously developers), given the number of \"gotchas\" in data processing, e.g., off-by-one issues in BED file loading."
}
]
},
{
"id": "14093",
"date": "01 Jun 2016",
"name": "Nicolas Servant",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present the InteractionSet package, that eases the manipulation of chromosome conformation data within the BioConductor/R framework. The InteractionSet package was designed to store direct interactions between two genomic loci. It also proposed a ContactMatrix class allowing to store the interaction counts as a Matrix format. One important point is its ability to be generic allowing the manipulation of any type of interaction data, such as ChIA-PET, Hi-C or 4C data. This work provides an interesting base for package development in this field and should therefore be of great use to the community.\n\nIn practice, the manuscript highlights the quality of the implementation and the optimization of this package. Dealing with Hi-C data can be challenging as the amount of data can be very large. Through this manuscript (Figure 1), it is clear that the authors propose an efficient strategy to manipulate such data.\nIn addition, the package is well documented with a quick start guide (vignette) and the description of each function. The new classes are based on existing S4 classes and methods and should therefore be easy to use for users familiar with the intervals manipulation in R. The package is already used as a dependency of other packages such as GenomicInteractions and diffHiC. Finally, it is compatible with other existing BioConductor packages such as the HiTC package.\n\nRegarding the manuscript itself, it clearly describes what the InteractionSet does. It is well written and easy to read.\n\nI only have a few minor comments that I hope will help the authors to improve the manuscript and/or the package.\n\nStoring direct interaction counts looks very interesting in practice. I'm just wondering how efficient is the GInteractions class in term of scalability and memory usage ? As an example, Rao et al recently generated Hi-C contact maps at a resolution of 5kb. This very high resolution dataset implies billion of Hi-C contacts. It would be interesting the know up to which resolution (or data throughput) the InteractionSet package is efficient, and/or which amount of RAM is require to deal with very large dataset.\n\nThe authors mentioned that the ContactMatrix class is compatible with matrix-based classes from the HiTC package. I therefore tried to convert a ContactMatrix object into a HTCexp object from HiTC (using the as() function) but It doesn't work. A note/example about that might be useful in the manual.\n\nThe package requires a recent version of R (>=3.3.0). It might be good to mention it somewhere.",
"responses": [
{
"c_id": "2006",
"date": "28 Jun 2016",
"name": "Aaron Lun",
"role": "Author Response",
"response": "Thanks for your comments Nicolas. Our responses are below: Storing direct interaction counts looks very interesting in practice. I'm just wondering how efficient is the GInteractions class in term of scalability and memory usage ? As an example, Rao et al recently generated Hi-C contact maps at a resolution of 5kb. This very high resolution dataset implies billion of Hi-C contacts. It would be interesting the know up to which resolution (or data throughput) the InteractionSet package is efficient, and/or which amount of RAM is require to deal with very large dataset. In several Hi-C analyses that we have performed (50 kbp resolution, ~1 billion reads), the size of the InteractionSet object is around 100MB to 1GB. This is well within the capacity of modern desktop machines, let alone high performance computing (HPC) facilities. For smaller bin sizes, the quadratic increase in the potential number of bin pairs is mitigated by the fact that a greater number of those bin pairs will be empty. We only store non-empty interactions in GInteractions/InteractionSet objects, which avoids a quadratic increase in memory requirements. (In practice, further savings can be made by filtering to remove low-abundance interactions.) That said, for very large data sets with read coverage across the entire interaction space, the memory requirements will increase dramatically at higher resolutions. This can be mitigated to some extent by only operating on a single chromosome (or pair of chromosomes) at any given time. However, if this is not possible (e.g., the downstream analysis requires all interactions), then HPC resources and 64-bit R may be required to handle the resulting objects. We feel that such requirements are mostly unavoidable, as the generation of large data sets requires concomitant effort in the computational analysis. The authors mentioned that the ContactMatrix class is compatible with matrix-based classes from the HiTC package. I therefore tried to convert a ContactMatrix object into a HTCexp object from HiTC (using the as() function) but It doesn't work. A note/example about that might be useful in the manual. Our concept of compatibility was based more on the class implementations and concepts, rather than through any explicit conversion. Specifically, both ContactMatrix and HTCexp use GRanges to represent the genomic coordinates of the row/column regions, and a Matrix class to represent store the intensity values across the interaction space. Thus, information extracted from an instance of a ContactMatrix class can directly be supplied to the HTCexp constructor: coords <- GRanges(\"chrA\", IRanges(1:10, 1:10)) x2 <- ContactMatrix(Matrix(1:100, 10, 10), coords, coords) # dummy object colnames(x2) <- rownames(x2) <- LETTERS[1:10] # dummy names HTCexp(as.matrix(x2), anchors(x2, \"column\"), anchors(x2, \"row\")) We are reluctant to implement this directly as an \"as\" method in the InteractionSet package, because we have tried to maintain a distinction between the low-level base classes in our package and the high-level analysis and visualization methods in other packages like HiTC, diffHic, GenomicInteractions, etc. Our hope is that developers who would like to use or be compatible with InteractionSet would write appropriate methods in their own packages to convert to/from classes as necessary. The package requires a recent version of R (>=3.3.0). It might be good to mention it somewhere. Done."
}
]
}
] | 1
|
https://f1000research.com/articles/5-950
|
https://f1000research.com/articles/5-1531/v1
|
28 Jun 16
|
{
"type": "Software Tool Article",
"title": "Robust de novo pathway enrichment with KeyPathwayMiner 5",
"authors": [
"Nicolas Alcaraz",
"Markus List",
"Martin Dissing-Hansen",
"Marc Rehmsmeier",
"Qihua Tan",
"Jan Mollenhauer",
"Henrik J. Ditzel",
"Jan Baumbach",
"Nicolas Alcaraz",
"Markus List",
"Martin Dissing-Hansen",
"Marc Rehmsmeier",
"Qihua Tan",
"Jan Mollenhauer",
"Henrik J. Ditzel"
],
"abstract": "Identifying functional modules or novel active pathways, recently termed de novo pathway enrichment, is a computational systems biology challenge that has gained much attention during the last decade. Given a large biological interaction network, KeyPathwayMiner extracts connected subnetworks that are enriched for differentially active entities from a series of molecular profiles encoded as binary indicator matrices. Since interaction networks constantly evolve, an important question is how robust the extracted results are when the network is modified. We enable users to study this effect through several network perturbation techniques and over a range of perturbation degrees. In addition, users may now provide a gold-standard set to determine how enriched extracted pathways are with relevant genes compared to randomized versions of the original network.",
"keywords": [
"Pathway enrichment",
"network analysis",
"data integration",
"algorithms",
"systems biology"
],
"content": "Introduction\n\nDe novo pathway enrichment methods have gained much attention during the last decade due to their potential to identify novel regulators and putative biomarkers from vast datasets in systems biology research. Given a biological interaction network, such as defined by BioGrid1, IntAct2 or I2D3, the main objective of de novo pathway enrichment is to extract connected subnetworks that are enriched for genes that are implicated in the phenotype of interest. This phenotype is dependent on the experiment and observed in one or several omics datasets, including, for instance, gene expression values, DNA methylation signals or single nucleotide variants. The common denominator of de novo pathway enrichment methods is that the resulting subnetworks are expected to include known pathways as well as novel pathways that have little overlap with annotated pathways found in curated databases. Existing approaches can be divided into the following categories: (i) aggregate score optimization methods, where the objective is to extract subnetworks that maximize a summary or statistical score of the individual gene scores, (ii) score propagation methods, where individual gene scores from the molecular profiles are propagated through the network, or adjusted to reflect also their connectivity in the network, (iii) module cover approaches, where the objective is to extract subnetworks containing genes that cover as many active cases/samples as possible, and (iv) cluster-based approaches. Methods that fall into categories (i), (ii) or (iv) rely heavily on the scoring function, which must be appropriate for the technology of the molecular profile being studied. In contrast, methods based on the module cover approach (iii) do not suffer from this issue, but leave it up to the user to find a sensible way to discern active from inactive genes. An overview of popular de novo pathway enrichment methods is shown in Table 1. We identify three issues common to existing de novo methods:\n\nThere is little consensus on what constitutes a novel pathway. It is up to the user to find method-specific parameters that lead to a satisfying solution. Choosing these parameters is often not intuitive and even small changes can lead to large variations in the results. Most methods provide little guidance on parameter selection, forcing users to rely on educated guesses, or to tediously re-run the method multiple times until the optimal parameters for a given analysis are found.\n\nIt has been demonstrated that for several methods results change significantly upon perturbations in the underlying networks4. This lack of robustness is an issue, since interaction databases are continuously evolving and it is unclear to what degree the results will change when a particular tool is applied with the exact same data to a newer version of the network.\n\nIn the rare cases where a ground truth or gold standard is available, the validation of de novo pathway enrichment results is not straightforward and, to our knowledge, not supported by any available method.\n\nAbbreviations: Cytoscape app (CA), standalone version/package (SA), desktop application (DA), web application (WA), web service (WS), visualization (VIZ), multi-omics (MO), robustness of the results upon network perturbation (RB), validation of the results using a gold standard upon network perturbation (VL).\n\nWe have previously developed KeyPathwayMiner, a de novo pathway enrichment tool following the module cover approach. Even though the parameters in KeyPathwayMiner are relatively intuitive, their selection becomes challenging for analyses involving several distinct omics datasets. To address this issue, we allow users to define a range (consisting of minimum, maximum and step size) for each parameter. The resulting grid search is fully automated and saves the user from going through tedious repetitions. While testing different parameters is more convenient in this way, it is still necessary to manually inspect the resulting subnetworks to select the optimal settings in a subjective fashion.\n\nHere, we present version 5 of KeyPathwayMiner, which is the first tool to provide a user-friendly way to systematically evaluate the quality and robustness of the results in de novo pathway enrichment. We achieve this by perturbing the input network to varying degrees. Depending on the research question, several perturbation techniques are available. To assess robustness of the results, the largest solution found in the perturbed network(s) is compared against the largest solution found in the unperturbed network. The size and variance of the overlap is illustrated for different user-controlled levels of perturbation and is an indicator for the robustness of the results. If a gold standard is available, an additional measure of quality is the overlap of the largest solution found in the unperturbed as well as in the perturbed network(s) with the gold standard. As an example application case, we apply KeyPathwayMiner to a gene expression dataset covering 38 Huntington’s disease patients and 32 healthy controls. We demonstrate the potential of network perturbation to help assessing the quality and robustness of the extracted results.\n\n\nMethods\n\nKeyPathwayMiner is implemented as a modular Java application centering on a core module that provides various de novo pathway enrichment strategies and methods for network perturbation analysis and plotting. A number of application modules have been implemented for different usage scenarios, including a standalone module, a web application module13, and a Cytoscape app module. The web application module KeyPathwayMinerWeb (http://keypathwayminer.compbio.sdu.dk), for instance, is primarily targeted at researchers with little to no experience in Cytoscape. No installation is necessary and convenience features, such as the mapping of identifiers or the conversion of p-value matrices to indicator matrices, are included. Web application developers may utilize a RESTful interface to integrate KeyPathwayMinerWeb seamlessly into their own applications. The standalone version is targeted at developers and data analysts who need more computational power than KeyPathwayMinerWeb offers and thus seek to incorporate KeyPathwayMiner directly in their own software implementation or in data analysis scripts, respectively. Finally, the Cytoscape app is the most powerful module, since it is also not limited with respect to the parameter range and computational power needed but also offers additional useful features such as combining OMICs datasets with a logical formula editor or the generation of plots using the JFreeChart (www.jfreechart.org) library.\n\nAfter installing the KeyPathwayMiner Cytoscape app via the app store, the user is expected to load an interaction network into Cytoscape. The KeyPathwayMiner tab can be found in the Control Panel and allows for one or more indicator matrices to be selected as input under the initial tab ‘Data’. These matrices can be derived from OMICs datasets such that samples correspond to columns and nodes (genes) to rows. Each entry in the matrix is either a ‘1’ indicating an active case in a node or ‘0’ otherwise. A typical example is a gene expression dataset in which a ‘1’ represents a differentially expressed gene. Another example could be a next-generation sequencing dataset where a ‘1’ indicates a single nucleotide polymorphism. Example files can be downloaded from the KeyPathwayMiner website at http://keypathwayminer.compbio.sdu.dk (Figure 1A). In the next tab, called ‘Links’, the user can customize how several datasets are combined for the analysis. Here, one can choose between ‘AND’ (a case is considered active if it is active in all datasets), ‘OR’ (a case is considered active if it is active in any of the datasets), or ‘CUSTOM’, which allows for connecting datasets in an interactive formula editor (Figure 1B). The tab ’Pos/Neg’ allows the user to define nodes that are always considered active (positive list) or that are ignored (negative list). In the ‘Run’ tab, it is finally possible to select the parameters for the KeyPathwayMiner run. Batch runs can also be performed by defining a range of values for K and L, such that users can conveniently run and assess the results for varying values of these parameters. (Figure 1C). KeyPathwayMiner relies on two easy-to-interpret parameters to control the size of the extracted subnetworks. The user can choose between a local as well as a global enrichment strategy. In the local strategy, INEs (Individual Node ExceptionS), a gene is considered active when it is active in all but L cases/samples. In addition, a parameter K allows KeyPathwayMiner to add additional inactive genes to extend the size of the solutions. We observe that INES has a tendency to prefer hub nodes, which is not always desirable. We therefore implemented the GLONE (GLobal Node Exceptions) strategy, where the parameter L is considered across all genes and fewer hub nodes are selected at the cost of run-time.\n\nThe user sets the analysis up as follows (omitting the ‘Pos/Neg’ tab, where nodes can be specified for inclusion or exclusion): (A) one or several dataset files are selected from the disk. (B) Several dataset files can be logically connected via a formula editor. (C) The run parameters are configured, most importantly the enrichment strategy (INES or GLONE), the search algorithm, the input network and the search parameters K and/or L, which can also be defined as a range. (D) Network perturbation settings used in robustness or validation runs.\n\nThe optimal values for K and L depend on the dataset14. KeyPathwayMiner allows users to define a range for both parameters to identify the best settings in a straight-forward fashion.\n\nUsers can choose between different methods to extract subnetworks: an exact (fixed parameter tractable), a greedy, and a heuristic (ant colony optimization) algorithm. For additional details regarding KeyPathwayMiner we refer to14–16.\n\nSeveral new features have been implemented in version 5 of KeyPathwayMiner and are described in the following.\n\nKeyPathwayMiner now enables users to study the robustness and validity of the extracted subnetworks through perturbation (Figure 1D). To this end, the user can choose from the following common strategies:\n\nNode label permutation: Pairs of nodes are selected arbitrarily and their node labels are swapped. This technique preserves the network structure exactly, but affects the local density of active genes in the network.\n\nDegree preserving rewiring: In this strategy first suggested by Maslov et al.17, two arbitrary edges are selected and their endpoints are swapped. As a result, the local network structure is actively changed while the global topological structure and the node degree distribution remain intact. With a large number of permutations this strategy leads to a randomized network.\n\nNode removal: In this strategy, a certain percentage of arbitrarily selected nodes are removed, thus simulating what results on a less complete network would look like. This is particularly interesting since interaction networks are continuously growing in size.\n\nEdge removal: In contrast to node removal, which affects network size, this strategy affects primarily the density of a network.\n\nTo assess the quality of the results, we consider the overlap of the largest solution between the various perturbed and the non-perturbed analyses. With an increasing degree of perturbation of the network, it can be expected that this overlap will decrease. Users can thus assess how robust the observed result is by considering the Jaccard similarity coefficient between the gene sets Sperturbed and Sunperturbed based on the gene sets extracted from the largest solution found using the perturbed and non-perturbed networks, respectively:\n\nSimilarly, the comparison of the overlap of the largest solution of the perturbed as well as the non-perturbed analyses and a gold standard SGold can be used as a quality metric:\n\nAs a convenience feature, we now allow users to select L, which allows users to define the number of case exceptions allowed in a solution, to be defined as a percentage of the total number of cases. This is particularly advantageous in the case of multiple datasets, where the L parameter (range) can now be selected once for all datasets in spite of differences in the number of cases between them.\n\nThe INES model extracts subnetworks with up to K exception nodes that are not active or differentially expressed (as defined by the L parameter). In many cases, these exception nodes are central in the pathway, i.e. they connect (groups of) active genes. However, if the K parameter is too large, some of these exception nodes are simply added to the periphery of the subnetwork to increase the size of the solution (Figure 2). As a result, the top solutions of a KeyPathwayMiner run would sometimes consist of overlapping subnetworks that only differ in these border exception nodes (BENs). BENs can now be removed in an optional filtering step, which will lead to more diverse solutions.\n\nRemoving BENs (red) would not disconnect regions containing no exception nodes (white). In contrast, removing non-BEN exception nodes (dark grey) would create two disjoint subnetworks containing non-exception nodes.\n\nBENs are removed as shown in Algorithm 1, which has worst-case running time of O((|V| + |E|) ∗ K).\n\n\nUse cases\n\nWe tested the usability of the new KeyPathwayMiner features with a gene expression dataset consisting of tissue samples from the caudate nucleus region of the brain18 taken from 38 patients suffering from Huntington’s disease (HD) and from 32 healthy patients in the control group. While it is known that huntingtin protein plays a major role in the development of the disease, the corresponding gene is not differentially expressed in approximately 40% of the patients. Hence, it will not be found in an analysis focused on identifying differentially expressed genes. However, it can be expected that protein-protein interaction (PPI) partners of the huntingtin protein are differentially expressed on the transcript level, thus posing an ideal test case for KeyPathwayMiner, which can identify subnetworks with huntingtin as an exception node. The Human Protein Reference Database (HPRD version 9, http://www.hprd.org/download)19 was used as the interaction network. To produce an indicator matrix for down-regulation, a one-tailed t-test was used to compute p-values for each gene and patient in the disease group vs all patients in the control group. Afterwards, a p-value cutoff of 0.05 was selected to set a 1 (significant) or 0 (not significant) in the indicator matrix for down-regulation (file available at http://keypathwayminer.compbio.sdu.dk/downloads/matrix-hd-down.dat).\n\nIn a first use case we performed a robustness analysis with KeyPathwayMiner (INES model, greedy algorithm) by fixing parameters to L = 20% (8 out of 32) of the cases and K = 5 exception nodes. In other words, we searched for maximal connected subgraphs containing at most 5 nodes with no more than 20% of cases in which the gene represented by the network node is not down-regulated.\n\n\n\nInput: Graph G(V, E), Exception Nodes Ve ⊂ V\n\nOutput: Subgraph G′ ⊆ G without BENs\n\nG′ := G ;\n\nwhile Ve≠0 do\n\nVben:=0 ;\n\nv := select and remove random node from Ve ;\n\nEv := edges incident to v ;\n\nGtemp := G′(V \\ {v}, E \\ Ev) ;\n\nC := connected components of Gtemp ;\n\ns := 0 ;\n\nforeach c ∈ C do\n\nif V(c) ∩ Ve == V(c) then\n\nVben := Vben ∪ V(c)\n\nelse\n\ns := s + 1 ;\n\nend\n\nend\n\nif s == 1 then\n\nVben := Vben ∪ {v}\n\nend\n\nVe := Ve \\ Vben ;\n\nEben := edges incident to Vben in G′ ;\n\nG′ := G′(V \\ Vben, E \\ Eben) ;\n\nend\n\nreturn G′ ;\n\nIn a typical robustness scenario, we wanted to test how the solutions change when a certain percentage of the edges in the graph are removed randomly. We thus selected \"edge removal\" as the perturbation technique with perturbation levels selected to range from 10% to 50% in increments of 10%. For each perturbation level, we created 10 randomly perturbed networks and executed KeyPathwayMiner with identical settings as in the original run.\n\nAs one would expect, removing a certain percentage of edges reduced the overlap with the results from the original network. However, even after removing 50% of the edges (Figure 3), a moderate Jaccard index overlap of 0.45 was observed.\n\nFor each perturbation level, 10 different networks were generated randomly and submitted to KeyPathwayMiner for analysis. (Parameters: INES, greedy, K = 5, L = 20%)\n\nIn a second use case, we give an example of a validation run. To this end, we compiled a gold standard gene set consisting of HD relevant genes (file available at http://keypathwayminer.compbio.sdu.dk/downloads/htt-relevant.txt) from the KEGG20,21 HD and calcium signaling pathways. Calcium signaling has been suggested to have an important role in the development of HD22.\n\nIn this scenario we wanted to see how solutions overlap with gold standard gene sets when randomly shuffling the node labels. In addition to the indicator matrix for down-regulation, we also aimed at finding solutions containing up-regulated genes. Hence we produced an additional indicator matrix for up-regulation (file available at http://keypathwayminer.compbio.sdu.dk/downloads/matrix-hd-up.dat) and connected them both with an ‘OR’ operator. We set a common L = 15% for both sets together with K = 5. KeyPathwayMiner (INES model, greedy algorithm) thus searched for pathways containing at most five genes with at most L = 15% genes that are not differentially regulated. The perturbation technique chosen was \"node label permutation\". The results show that even when permuting 80% of the node labels the overlap with the gold standard set remains relatively stable. As expected, we can see a significant drop when all labels are permuted (Figure 4).\n\nFor each perturbation degree, 10 graphs were generated. (Parameters: INES, greedy, K = 5, L = 15%). Note that partial network perturbation and subsequent comparison with gold standard sets has limited meaning. Biological interaction networks are scale-free, i.e. robust to random perturbations. Of major importance for this effect are a small number of hub nodes. KeyPathwayMiner is able to recover subnetworks containing relevant genes connected to such hubs unless the hubs themselves are affected by the perturbation. This, however, is only the case when 100% of the network is perturbed (a randomized, true null model), explaining the performance drop we observe for this degree of perturbation.\n\n\nSummary\n\nDe novo pathway enrichment is a powerful method for the analysis of one or several types of molecular profiles. In contrast to widely used gene set enrichment methods such as GSEA23, this methodology is not limited to existing knowledge but suitable to uncover new functional modules. Results are extracted using large biological interaction networks, which are incomplete and continuously evolve. It is typically unclear how future updates leading to an interaction network of higher quality will affect the currently obtained results. KeyPathwayMiner 5 enables users to study the robustness of their results by allowing them to introduce artificial noise into the underlying interaction networks. Moreover, an existing gold standard can be used to test how well the optimal solution can be recovered on these perturbed networks.\n\n\nData availability\n\nF1000Research: Dataset 1. Use case data of de novo pathway enrichment with KeyPathwayMiner 5, 10.5256/f1000research.9054.d12687124\n\n\nSoftware availability\n\n1. Software available from: http://apps.cytoscape.org/apps/keypathwayminer\n\n2. Latest source code: https://github.com/baumbachlab/keypathwayminer-cytoscape3 https://github.com/baumbachlab/keypathwayminer-cytoscape3/archive/5.0.tar.gz (KeyPathwayMiner Cytoscape app source code) https://github.com/baumbachlab/keypathwayminer-core/archive/5.0.tar.gz (KeyPathwayMiner core library source code)\n\n3. Archived source code as at time of publication: Zenodo, Source codes de novo pathway enrichment with KeyPathwayMiner, doi: 10.5281/zenodo5573425\n\n4. License: GPL v3",
"appendix": "Author contributions\n\n\n\nJB, MR, NA, and ML specified the new features and plots that were implemented by MDH and NA. NA, ML and JB wrote the manuscript. All authors contributed to testing the software and have read and approved the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Lundbeckfonden grant for the NanoCAN Center of Excellence in Nanomedicine, the Region Syddanmarks ph.d.-pulje and Forskningspulje, the Fonden Til Lægevidenskabens Fremme, by the DAWN-2020 project financed by Rektorspuljen SDU2020 program, the MIO project of the OUH Frontlinjepuljen, the Bioinformatics part of NEXT – National Experimental Therapy Partnership funded by the Innovation Fund Denmark, as well as the VILLUM foundation by a Blokstipendiet. NA would like to acknowledge el Consejo Nacional de Ciencia y Tecnología (CONACyT) from Mexico for their financial support.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nChatr-Aryamontri A, Breitkreutz BJ, Oughtred R, et al.: The BioGRID interaction database: 2015 update. Nucleic Acids Res. 2015; 43(Database issue): D470–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrchard S, Ammari M, Aranda B, et al.: The MIntAct project--IntAct as a common curation platform for 11 molecular interaction databases. Nucleic Acids Res. 2014; 42(Database issue): D358–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrown KR, Jurisica I: Unequal evolutionary conservation of human protein interactions in interologous networks. Genome Biol. 2007; 8(5): R95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeisser D, Brunkhorst S, Dandekar T, et al.: Robustness and accuracy of functional modules in integrated network analysis. Bioinformatics. 2012; 28(14): 1887–1894. PubMed Abstract | Publisher Full Text\n\nBeisser D, Klau GW, Dandekar T, et al.: BioNet: an R-Package for the functional analysis of biological networks. Bioinformatics. 2010; 26(8): 1129–30. PubMed Abstract | Publisher Full Text\n\nBreitling R, Amtmann A, Herzyk P: Graph-based iterative Group Analysis enhances microarray interpretation. BMC Bioinformatics. 2004; 5: 100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNacu S, Critchley-Thorne R, Lee P, et al.: Gene expression network analysis and applications to immunology. Bioinformatics. 2007; 23(7): 850–858. PubMed Abstract | Publisher Full Text\n\nVandin F, Upfal E, Raphael BJ: Algorithms for detecting significantly mutated pathways in cancer. J Comput Biol. 2011; 18(3): 507–22. PubMed Abstract | Publisher Full Text\n\nIdeker T, Ozier O, Schwikowski B, et al.: Discovering regulatory and signalling circuits in molecular interaction networks. Bioinformatics. 2002; 18(Suppl 1): S233–S240. PubMed Abstract | Publisher Full Text\n\nUlitsky I, Shamir R: Identification of functional modules using network topology and high-throughput data. BMC Syst Biol. 2007; 1: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChuang HY, Lee E, Liu YT, et al.: Network-based classification of breast cancer metastasis. Mol Syst Biol. 2007; 3: 140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQiu YQ, Zhang S, Zhang XS, et al.: Detecting disease associated modules and prioritizing active genes based on high throughput data. BMC Bioinformatics. 2010; 11: 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nList M, Alcaraz N, Dissing-Hansen M, et al.: KeyPathwayMinerWeb: online multi-omics network enrichment. Nucleic Acids Res. 2016; pii: gkw373. PubMed Abstract | Publisher Full Text\n\nAlcaraz N, Pauling J, Batra R, et al.: KeyPathwayMiner 4.0: condition-specific pathway analysis by combining multiple omics studies and networks with Cytoscape. BMC Syst Biol. 2014; 8: 99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlcaraz N, Kücük H, Weile J, et al.: KeyPathwayMiner: Detecting Case-Specific Biological Pathways Using Expression Data. Internet Mathematics. 2011; 7(4): 299–313. Publisher Full Text\n\nAlcaraz N, Friedrich T, Kötzing T, et al.: Efficient key pathway mining: combining networks and OMICS data. Integr Biol (Camb). 2012; 4(7): 756–64. PubMed Abstract | Publisher Full Text\n\nMaslov S, Sneppen K: Specificity and stability in topology of protein networks. Science. 2002; 296(5569): 910–913. PubMed Abstract | Publisher Full Text\n\nHodges A, Strand AD, Aragaki AK, et al.: Regional and cellular gene expression changes in human Huntington’s disease brain. Hum Mol Genet. 2006; 15(6): 965–977. PubMed Abstract | Publisher Full Text\n\nPrasad TS, Kandasamy K, Pandey A: Human Protein Reference Database and Human Proteinpedia as discovery tools for systems biology. Methods Mol Biol. 2009; 577: 67–79. PubMed Abstract | Publisher Full Text\n\nKanehisa M, Goto S: KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000; 28(1): 27–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanehisa M, Sato Y, Kawashima M, et al.: Kegg as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016; 44(D1): D457–D462. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRockabrand E, Slepko N, Pantalone A, et al.: The first 17 amino acids of Huntingtin modulate its sub-cellular localization, aggregation and effects on calcium homeostasis. Hum Mol Genet. 2007; 16(1): 61–77. PubMed Abstract | Publisher Full Text\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlcaraz N, List M, Dissing-Hansen M, et al.: Dataset 1 in: Robust de novo pathway enrichment with KeyPathwayMiner 5. F1000Research. 2016. Data Source\n\nAlcaraz N, List M, Dissing-Hansen M, et al.: Source codes de novo pathway enrichment with KeyPathwayMiner. Zenodo. Data Source"
}
|
[
{
"id": "14664",
"date": "01 Jul 2016",
"name": "Alberto Calderone",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article presents an update to a previously published Cytoscape App by providing a general overview of other methods. The title, as well as the abstract give a good introduction to the article.\nThe authors give a good recap of other methods and possible approaches as well as presenting the updated app. The Cytoscape app presentation is clear and enough for the final user. Most importantly, the new feature introduced in V5 - i.e. Perturbation - is clearly described in a dedicated paragraph. \"Robust\" in the title is justified in the text and use cases were clear and doable. Border exception node is illustrated clearly as well.\nI could reproduce the examples given with no big problem. Overall, the article is well written.",
"responses": []
},
{
"id": "15263",
"date": "28 Jul 2016",
"name": "Mona Riemenschneider",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript describes a tool for robust de novo pathway enrichment. The software provides the great advantage to study effects of network perturbations thereby allow for the evaluation of quality and robustness of the results in de novo pathway enrichment. Thus, the authors address a relevant issue in network construction and pathway enrichment.\nThe rationale for the development of the tool is clearly stated. A use case to demonstrate the usability of KeyPathwayMiner 5 with varying parameters is described within the manuscript. The source code of KeyPathwayMiner 5 is available at github.com.\n\nMINOR COMMENTS:\nThe introduced tool KeyPathwayMiner 5 is a further development of KeyPathwayMiner x. A short overview of added features to all updated versions may be helpful for users to get an overview of the full function of KeyPathwayMiner 5. (provide in supplement)\n\nSeveral parameters must be set to run KeyPathwayMiner. Could an approximate recommendation for parameter values and settings be given for non-expert users?\n\nThe calculation of the Jaccard index is given in detail, however a short explanation of the graduation (high, moderate, low) of calculated values could be helpful to provide easy interpretation of results for all users.\n\nPlease check spelling throughout the manuscript: KeyPathwayMiner/KeyPathway Miner",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1531
|
https://f1000research.com/articles/5-1529/v1
|
28 Jun 16
|
{
"type": "Review",
"title": "Recent advances in molecular genetics of melanoma progression: implications for diagnosis and treatment",
"authors": [
"Iwei Yeh"
],
"abstract": "According to the multi-step carcinogenesis model of cancer, initiation results in a benign tumor and subsequent genetic alterations lead to tumor progression and the acquisition of the hallmarks of cancer. This article will review recent discoveries in our understanding of initiation and progression in melanocytic neoplasia and the impact on diagnostic dermatopathology.",
"keywords": [
"melanoma",
"dermatopathology",
"melanocytic neoplasia",
"BRAFV600E",
"BRAF",
"nevi",
"Spitz nevi"
],
"content": "Initiating oncogenes in melanocytic neoplasia\n\nIf an initiating oncogene causes tumor formation, it should be present clonally in benign neoplasms and occur in a mutually exclusive pattern with other initiating events. BRAFV600E satisfies these criteria in melanocytic neoplasia. Studies demonstrate that BRAF mutations are typically present in all or none of the cells within nevi and melanomas1,2. In a recent study of the genetic evolution of melanoma, sequencing of known oncogenes in melanoma and cancer did not reveal additional driver alterations in unequivocally benign nevi with BRAFV600E, additionally supporting the hypothesis that BRAFV600E mutation can initiate melanocytic nevi3.\n\nThe set of probable initiating oncogenes in melanocytic tumors includes activating point mutations in BRAF, NRAS, GNAQ, GNA11, and activating fusions of BRAF and the receptor tyrosine kinases (RTKs) ALK, ROS1, RET, MET, and NTRK1. These mutations have been identified in benign and malignant melanocytic tumors in a mutually exclusive pattern, e.g. only one of these MAPK-activating mutations will be present. While KIT mutations do not co-occur with probable initiating oncogenes, they have not been identified in benign melanocytic tumors and it is unclear when activating KIT mutations arise during melanoma progression.\n\nInitiating oncogenes may influence tumor phenotype. Different histopathologic subtypes of nevi demonstrate varying spectra of initiating mutations. Common acquired nevi harbor BRAFV600E mutations in ~85% of cases with activating NRAS mutations in 3–5%4. The majority of blue nevi harbor activating mutations in one of two highly homologous members of the G-alpha Q family, GNAQ (65%) and GNA11 (9%)5,6. Spitz nevi have the most diverse set of initiating mutations with activating HRAS mutations (22%) and activating fusions of BRAF (7%) and the RTKs ALK (12%), MET (2%), NTRK1 (12%), RET (4%), and ROS1 (25%)7–10.\n\nPerhaps the phenotypic differences between common acquired nevi and blue nevi are due to their distinct initiating oncogenes. If so, the diversity of initiating oncogenes in Spitz nevi and tumors could explain the phenotypic variability and the diagnostic challenges of this class of tumors. Initial studies suggest that fusions of specific RTKs may result in specific histopathologic features. Specifically, Spitz tumors with ALK fusions commonly have distinctive vertically oriented plexiform nests of fusiform melanocytes11,12. Classification of Spitz tumors by the category of initiating oncogene may result in more refined histopathologic diagnostic criteria. One caveat is the diversity within a given class of fusion kinases. Structural rearrangements lead to oncogenic RTK fusion genes because the N-terminal fusion partner replaces the regulatory portion of the RTK. Without the regulatory domain, the kinase domain is constitutively active. Early findings indicate that activating fusions of the same RTK may be highly diverse in melanocytic tumors with a broad range of N-terminal partners fused to variable portions of the RTK8–10. The N-terminal partner influences expression, localization, and dimerization of the fusion kinase, all features expected to impact oncogenic signaling and thus potentially tumor phenotype.\n\n\nProgression events\n\nThe accumulation of oncogenic events in addition to an initiating event leads to melanoma. Owing to the high number of mutations observed in melanoma, distinguishing driver from passenger events is difficult and requires functional validation. Through large-scale sequencing studies, many progression events have been nominated in melanoma, but the functional consequences of most of them remain to be determined13–15. Understanding how combinations of oncogenic mutations interact and predict biologic behavior is an area of active investigation.\n\nFirst identified in 2012 in familial and sporadic cutaneous melanoma, TERT promoter mutations result in a de novo E26 transformation-specific (ETS) factor binding site and increased TERT expression16–18. TERT is the enzymatic subunit of telomerase, and elevated telomerase activity prevents critical telomere shortening with cell division and bypasses replicative senescence. TERT promoter mutations are associated with worse prognosis in non-acral cutaneous melanoma and Spitzoid melanoma19,20.\n\nA recent study of melanoma progression characterized various portions of melanocytic tumors that contained benign, intermediate, and malignant areas (melanoma arising within a nevus). TERT promoter mutations were identified in several “likely benign” intermediate melanocytic tumors and melanomas3. The presence of TERT promoter mutations in combination with either BRAF or NRAS activating mutations in “likely benign” intermediate tumors suggests that these combinations of oncogenic mutations are not sufficient for malignant transformation. This finding demonstrates that an intermediate category of melanocytic neoplasia exists and corresponds with existing histopathologic classifications.\n\nA recent study identified a novel mechanism that results in translation of the kinase portion of an RTK without its corresponding regulatory domain. A novel transcript of ALK transcribed from an alternative transcription initiation (ATI) site in intron 19 of the full-length isoform of ALK encodes the kinase domain of ALK without the extracellular or transmembrane regions. Present in ~3–11% of melanomas, ALKATI is not associated with DNA sequence alterations of ALK. Rather, it appears that the expression of ALKATI occurs due to epigenetic modification. In vitro, ALKATI constitutively activates MAPK, AKT, and STAT3 signaling and is inhibited by small molecule ALK inhibitors. While the signaling output of ALKATI is similar to that of ALK fusions, ALKATI is seen in melanomas with and without activating BRAF and NRAS mutations, indicating that it is not an initiating event21.\n\nBiallelic BAP1 loss can in many cases be distinctly identified by histopathologic examination. BAP1 is a histone deubiquitinase that functions as a tumor suppressor. It is recurrently inactivated in uveal melanoma22. Germline loss-of-function variants increase the risk of melanoma, renal cell carcinoma, mesothelioma, and other cancers23,24. The distinctive cutaneous melanocytic tumors in patients with BAP1 germline mutations are characterized by dermal epithelioid melanocytes with abundant eosinophilic cytoplasm and variably enlarged, pleomorphic, and eccentrically placed nuclei, often in a background of lymphocytic inflammation. These neoplasms harbor activating BRAF or NRAS mutations in addition to biallelic loss of BAP1 and an adjacent common acquired nevus is often appreciated25,26. These findings are consistent with clonal expansion of a neoplastic melanocyte in a common acquired nevus (with BRAF or NRAS activating mutation) after biallelic loss of BAP1. Based on their cytomorphology, epithelioid tumors with BAP1 loss were historically classified as atypical Spitz tumors or halo Spitz nevi, both considered to have negligible to low malignant potential.\n\nEpithelioid tumors with BAP1 loss (or Wiesner nevi) are distinct from other genetic categories of Spitz nevi in that three oncogenic mutations have occurred (activating BRAF or NRAS mutation and two hits to BAP1) in contrast to Spitz nevi with HRAS mutation or kinase fusions. Their characteristic cytomorphology is due to a progression event (loss of BAP1) rather than a direct effect of the initiating oncogene, as is hypothesized for other Spitz nevi. Thus, there is an argument to be made to cleave these tumors from the Spitz progression series and add them as a subtype of intermediate tumor on the BRAF/NRAS progression series.\n\nEarly observations indicate that BAP1 loss in combination with BRAF or NRAS mutation gives rise to a low-risk melanocytic tumor (a topic worthy of further investigation). In contrast, BAP1 loss in combination with GNAQ or GNA11 has not been identified in low-risk melanocytic tumors but occurs in uveal melanoma and melanoma arising in blue nevi (MABN)22,27. Loss of BAP1 is associated with poor prognosis in uveal melanoma28. Thus, the contribution of BAP1 loss to malignant transformation in melanoma appears to differ depending on the initiating oncogene. Our models of melanoma progression will need to accommodate this complexity.\n\n\nGenomic reflections of aberrant cellular processes\n\nArm-level and whole chromosome gains and losses, as well as focal amplifications and deletions of the genome, are frequent in melanoma and uncommon in nevi. Copy number aberrations (CNAs), particularly when multiple, may reflect previous or ongoing genomic instability. Genomic instability can result from multiple disrupted biologic processes (oncogene-induced replicative stress, defective DNA damage response, or impaired cell cycle checkpoints).\n\nThe overrepresentation of specific copy number alterations in melanoma indicates selective advantage for specific CNAs (i.e. loss of CDKN2A or amplification of CCND1) and a role in tumor progression. Melanomas arising on chronically sun-damaged skin, non-chronically sun-damaged skin, acral glabrous skin, and mucosal epithelium have different patterns of CNAs, suggesting different causes of genomic instability and/or different pathways of genetic evolution29. Not all types of melanoma demonstrate a high frequency of CNAs: for example, desmoplastic melanomas have few CNAs and a high number of single base substitutions30.\n\nWhile CNAs may reflect genomic instability, they can also result from stochastic events in the absence of a long-term cellular state of global genomic instability. These events include double-stranded DNA breaks and catastrophic events that lead to complex genomic rearrangements, such as chromothripsis31. In benign or low-grade melanocytic tumors, such a chance event is thought to give rise to CNAs that lead to selective advantage and selection. Gain of chromosome 11p is often observed in HRAS mutant Spitz nevi7. Monosomy 3 or focal loss including 3p21 is often observed in epithelioid tumors with biallelic loss of BAP123,25. Identification of these isolated CNAs in the context of a tumor with the expected histopathologic characteristics does not lead to a diagnosis of melanoma. Copy number transitions within kinases may indicate a kinase fusion. Often times we observe probable “passenger” structural variants in the vicinity of kinase fusions (for example, the reciprocal fusion junction). The clinical significance of varying patterns of copy number alterations seen in association with kinase fusions remains to be determined.\n\n\nMolecular assessment for diagnosis\n\nAssessment of copy number status has been used to supplement the histopathologic assessment of diagnostically challenging melanocytic tumors for over a decade. Array comparative genomic hybridization (aCGH) and fluorescence in situ hybridization (FISH) are in routine use by several diagnostic laboratories. One of the first FISH tests proposed for melanoma diagnosis employs four probes, assessing for gains of CCND1 and absolute or relative gain of 6p or loss of 6q as compared to centromere 632,33. Additional FISH probes have been proposed for specific subtypes of melanocytic tumors (9p21 to assess for homozygous CDKN2A deletion in spitzoid tumors and 8q24 MYC gain to improve sensitivity in nevoid melanomas)34,35.\n\naCGH gives a broader assessment of copy number status but is less sensitive in the setting of low tumor purity and for subclonal CNAs and also requires more tissue than FISH. The patterns of CNAs are varied and the significance of a limited number of CNAs that are not common in melanoma remains to be determined. The copy number profile can provide clues to oncogenic alterations. For example, KIT amplification is often associated with KIT mutation, and copy number transitions in kinases with relative gain of the kinase portion of the gene may indicate an activating kinase fusion.\n\nInitial studies highlight the promise of assessment of combinations of genetic alterations and expression profiles using multiplex analysis of DNA or RNA in the diagnosis of melanocytic neoplasia3,36. Additional studies with clinical follow-up and stratification by histopathologic and genetic subtype will inform how best to integrate these complex tests into current clinical practice.\n\n\nMolecular assessment for treatment selection\n\nCurrently, the two major approaches to the treatment of metastatic melanoma are immunotherapy and molecularly targeted therapies, and there are studies underway to evaluate combination regimens. The checkpoint inhibitors ipilimumab (anti CTLA-4 antibody), nivolumab, and pembrolizumab (anti PD-1 antibodies) result in objective responses in 10–40% of patients and an overall survival benefit37–40. PD-L1 expression correlates with response to anti PD-1 antibodies, and the combination of nivolumab and ipilimumab improves response rates in PD-L1-negative tumors. As the side effect profile of checkpoint inhibitors is not insignificant, work is currently ongoing to identify which patients will benefit from these treatments. In non-small-cell lung cancer, a higher mutation burden (likely a proxy for increased neoantigens) is associated with improved response to immunotherapy41. Estimation of mutation burden, neoantigen expression, or expression profile may refine therapy selection for metastatic melanoma in the near future42.\n\nTargeted therapy of BRAFV600E mutant melanoma with inhibitors of mutant BRAF is currently part of the standard of care. Combination with MEK inhibitors improves outcomes43,44. Approximately 50% of metastatic melanomas harbor a BRAF mutation, ~25% harbor an activating NRAS mutation, and 3–5% harbor an activating KIT mutation. Inhibitors of NRAS are currently unavailable, but initial clinical trials of MEK inhibitors in NRAS mutant melanomas show some efficacy45. Dramatic responses to KIT inhibitors such as imatinib and nilotinib have been observed in patients with KIT mutant melanoma46–50.\n\nIn a minority of cutaneous melanoma patients, an activating mutation in BRAF, NRAS, or KIT is not identified. In these patients, testing for a kinase fusion may yield a potential therapeutic target. In case reports of patients with BRAF fusion melanoma, responses to sorafenib and trametinib were observed8,51,52. Treatment of other solid tumors with RTK fusions similar to those observed in melanoma provides clinical benefit as exemplified by ALK inhibition in lung cancer with ALK fusions53,54. Clinical studies are needed to assess the efficacy of kinase inhibitors for kinase fusion melanoma.\n\nThere are an increasing number of diagnostic modalities available for the detection of actionable and potentially actionable genetic alterations. Considerations for selecting specific assays include cost, turn-around time, comprehensiveness for actionable alterations (a moving target), and specimen requirements. For point mutations, immunohistochemistry (VE1 for BRAFV600E and SP174 for NRASQ61R) and allele-specific real-time polymerase chain reaction (RT-PCR) assays (cobas® 4800 BRAF V600) provide quick, highly sensitive, and easy-to-interpret assessment for a narrow spectrum of mutations55–57. Sanger sequencing has been traditionally used for the detection of hotspot mutations in oncogenes and can detect mutations within the assayed region (i.e. BRAF exon 15). One limitation of Sanger sequencing is a limit of detection of ~10–20% minor allele frequency (corresponding to 20–40% tumor fraction in a heterozygous sample), resulting in decreased sensitivity for samples with low tumor fraction. Next-generation multiplex sequencing is being increasingly adopted as a way to perform multiplex testing of oncogenes with a lower limit of detection owing to the ability to sequence individual DNA molecules. Next-generation sequencing (NGS) cancer testing platforms typically assess a panel of oncogenes that are of interest in many types of cancer. By broadening the regions of the genome assayed, these panels may detect alterations that are actionable in other cancer types and rare in melanoma. These assays can also detect CNAs.\n\nOne can take advantage of the mutual exclusivity of actionable alterations and their prevalence in melanoma to perform stratified testing of a tumor sample. Given the high rate of BRAF V600 mutations, V600E-specific testing (immunohistochemistry or real-time based assay) or BRAF exon 15 testing (Sanger) followed by a test for a broader panel of oncogenes (including NRAS and KIT) if a BRAF mutation is not detected could optimize cost and turn-around time for melanoma patients, depending on testing strategies employed.\n\nIdentification of kinase fusions requires different approaches than the detection of oncogenic hotspot mutations, as the genomic breakpoints usually occur in intronic regions that span a much larger portion of the genome than hotspot coding mutations. Detection of fusion transcripts by RT-PCR is highly sensitive (i.e. BCR-ABL in chronic myelogenous leukemia), but RT-PCR is not practical for detecting the broad spectrum of kinase fusions that occur in melanoma. Immunohistochemistry to assess the expression of the kinase domain of ALK, ROS1, NTRK1, and MET appear to be highly sensitive for detecting fusion kinases but with varying specificity. The lack of specificity can be due to basal expression of the kinase in melanocytes (NTRK1 and MET) and alterative oncogenic mechanisms that lead to expression of the kinase domain (ALKATI). Hybrid-capture-based NGS DNA assays can detect structural rearrangements that lead to oncogenic fusions by sequencing the introns in which the breakpoints occur and can be multiplexed with detection of other melanoma oncogenes, but this method has limited sensitivity due to repetitive regions within introns and the technical difficulty of identifying structural rearrangements from short-read sequencing. Multiplex RNA-based methods are more sensitive. FISH break-apart probes are also available for fusion detection.\n\n\nFuture directions\n\nThe rapid pace of technologic development has led to a remarkable expansion of our understanding of the genetic progression of cancer and melanoma. Translation of these findings into the clinic is exceeding at a rapid pace. As always, we are treating patients with the best information we have on hand while pushing for additional studies to support our current best practices in diagnosis and treatment. Refining our understanding and models of genetic progression will help us develop the best clinical and biologic hypotheses to direct future investigation.",
"appendix": "Competing interests\n\n\n\nThe author declares that she has no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nYeh I, von Deimling A, Bastian BC: Clonal BRAF mutations in melanocytic nevi and initiating role of BRAF in melanocytic neoplasia. J Natl Cancer Inst. 2013; 105(12): 917–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRiveiro-Falkenbach E, Villanueva CA, Garrido MC, et al.: Intra- and Inter-Tumoral Homogeneity of BRAFV600E Mutations in Melanoma Tumors. J Invest Dermatol. 2015; 135(12): 3078–85. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nShain AH, Yeh I, Kovalyshyn I, et al.: The Genetic Evolution of Melanoma from Precursor Lesions. N Engl J Med. 2015; 373(20): 1926–36. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPollock PM, Harper UL, Hansen KS, et al.: High frequency of BRAF mutations in nevi. Nat Genet. 2003; 33(1): 19–20. PubMed Abstract | Publisher Full Text\n\nVan Raamsdonk CD, Bezrookove V, Green G, et al.: Frequent somatic mutations of GNAQ in uveal melanoma and blue naevi. Nature. 2009; 457(7229): 599–602. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nvan Raamsdonk CD, Griewank KG, Crosby MB, et al.: Mutations in GNA11 in uveal melanoma. N Engl J Med. 2010; 363(23): 2191–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBastian BC, LeBoit PE, Pinkel D: Mutations and copy number increase of HRAS in Spitz nevi with distinctive histopathological features. Am J Pathol. 2000; 157(3): 967–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBotton T, Yeh I, Nelson T, et al.: Recurrent BRAF kinase fusions in melanocytic tumors offer an opportunity for targeted therapy. Pigment Cell Melanoma Res. 2013; 26(6): 845–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiesner T, He J, Yelensky R, et al.: Kinase fusions are frequent in Spitz tumours and spitzoid melanomas. Nat Commun. 2014; 5: 3116. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nYeh I, Botton T, Talevich E, et al.: Activating MET kinase rearrangements in melanoma and Spitz tumours. Nat Commun. 2015; 6: 7174. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBusam KJ, Kutzner H, Cerroni L, et al.: Clinical and pathologic findings of Spitz nevi and atypical Spitz tumors with ALK fusions. Am J Surg Pathol. 2014; 38(7): 925–33. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nYeh I, de la Fouchardiere A, Pissaloux D, et al.: Clinical, histopathologic, and genomic features of Spitz tumors with ALK fusions. Am J Surg Pathol. 2015; 39(5): 581–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrauthammer M, Kong Y, Ha BH, et al.: Exome sequencing identifies recurrent somatic RAC1 mutations in melanoma. Nat Genet. 2012; 44(9): 1006–14. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nHodis E, Watson IR, Kryukov GV, et al.: A landscape of driver mutations in melanoma. Cell. 2012; 150(2): 251–63. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCancer Genome Atlas Network: Genomic Classification of Cutaneous Melanoma. Cell. 2015; 161(7): 1681–96. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nHuang FW, Hodis E, Xu MJ, et al.: Highly recurrent TERT promoter mutations in human melanoma. Science. 2013; 339(6122): 957–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nHorn S, Figl A, Rachakonda PS, et al.: TERT promoter mutations in familial and sporadic melanoma. Science. 2013; 339(6122): 959–61. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBell RJ, Rube HT, Kreig A, et al.: Cancer. The transcription factor GABP selectively binds and activates the mutant TERT promoter in cancer. Science. 2015; 348(6238): 1036–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nGriewank KG, Murali R, Puig-Butille JA, et al.: TERT promoter mutation status as an independent prognostic factor in cutaneous melanoma. J Natl Cancer Inst. 2014; 106(9): pii: dju246. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nLee S, Barnhill RL, Dummer R, et al.: TERT Promoter Mutations Are Predictive of Aggressive Clinical Behavior in Patients with Spitzoid Melanocytic Neoplasms. Sci Rep. 2015; 5: 11200. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBusam KJ, Villain RE, Lum T, et al.: Primary and Metastatic Cutaneous Melanomas Express ALK Through Alternative Transcriptional Initiation. Am J Surg Pathol. 2016; 40(6): 786–95. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nHarbour JW, Onken MD, Roberson ED, et al.: Frequent mutation of BAP1 in metastasizing uveal melanomas. Science. 2010; 330(6009): 1410–3. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWiesner T, Obenauf AC, Murali R, et al.: Germline mutations in BAP1 predispose to melanocytic tumors. Nat Genet. 2011; 43(10): 1018–21. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nOhar JA, Cheung M, Talarchek J, et al.: Germline BAP1 Mutational Landscape of Asbestos-Exposed Malignant Mesothelioma Patients with Family History of Cancer. Cancer Res. 2016; 76(2): 206–15. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nYeh I, Mully TW, Wiesner T, et al.: Ambiguous melanocytic tumors with loss of 3p21. Am J Surg Pathol. 2014; 38(8): 1088–95. PubMed Abstract | Free Full Text\n\nWiesner T, Murali R, Fried I, et al.: A distinct subset of atypical Spitz tumors is characterized by BRAF mutation and loss of BAP1 expression. Am J Surg Pathol. 2012; 36(6): 818–30. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCosta S, Byrne M, Pissaloux D, et al.: Melanomas Associated With Blue Nevi or Mimicking Cellular Blue Nevi: Clinical, Pathologic, and Molecular Study of 11 Cases Displaying a High Frequency of GNA11 Mutations, BAP1 Expression Loss, and a Predilection for the Scalp. Am J Surg Pathol. 2016; 40(3): 368–77. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nKalirai H, Dodson A, Faqir S, et al.: Lack of BAP1 protein expression in uveal melanoma is associated with increased metastatic risk and has utility in routine prognostic testing. Br J Cancer. 2014; 111(7): 1373–80. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCurtin JA, Fridlyand J, Kageshita T, et al.: Distinct sets of genetic alterations in melanoma. N Engl J Med. 2005; 353(20): 2135–47. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nShain AH, Garrido M, Botton T, et al.: Exome sequencing of desmoplastic melanoma identifies recurrent NFKBIE promoter mutations and diverse activating mutations in the MAPK pathway. Nat Genet. 2015; 47(10): 1194–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nForment JV, Kaidi A, Jackson SP: Chromothripsis and cancer: causes and consequences of chromosome shattering. Nat Rev Cancer. 2012; 12(10): 663–70. PubMed Abstract | Publisher Full Text\n\nGerami P, Jewell SS, Morrison LE, et al.: Fluorescence in situ hybridization (FISH) as an ancillary diagnostic tool in the diagnosis of melanoma. Am J Surg Pathol. 2009; 33(8): 1146–56. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMarch J, Hand M, Truong A, et al.: Practical application of new technologies for melanoma diagnosis: Part II. Molecular approaches. J Am Acad Dermatol. 2015; 72(6): 943–58; quiz 959–60. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nGerami P, Scolyer RA, Xu X, et al.: Risk assessment for atypical spitzoid melanocytic neoplasms using FISH to identify chromosomal copy number aberrations. Am J Surg Pathol. 2013; 37(5): 676–84. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nYélamos O, Busam KJ, Lee C, et al.: Morphologic clues and utility of fluorescence in situ hybridization for the diagnosis of nevoid melanoma. J Cutan Pathol. 2015; 42(11): 796–806. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nClarke LE, Warf MB, Flake DD 2nd, et al.: Clinical validation of a gene expression signature that differentiates benign nevi from malignant melanoma. J Cutan Pathol. 2015; 42(4): 244–52. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nLarkin J, Chiarion-Sileni V, Gonzalez R, et al.: Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. N Engl J Med. 2015; 373(1): 23–34. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nHodi FS, O'Day SJ, McDermott DF, et al.: Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010; 363(8): 711–23. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nRobert C, Long GV, Brady B, et al.: Nivolumab in previously untreated melanoma without BRAF mutation. N Engl J Med. 2015; 372(4): 320–30. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nRobert C, Schachter J, Long GV, et al.: Pembrolizumab versus Ipilimumab in Advanced Melanoma. N Engl J Med. 2015; 372(26): 2521–32. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nRizvi NA, Hellmann MD, Snyder A, et al.: Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science. 2015; 348(6230): 124–8. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nSpranger S, Bao R, Gajewski TF: Melanoma-intrinsic β-catenin signalling prevents anti-tumour immunity. Nature. 2015; 523(7559): 231–5. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nLong GV, Weber JS, Infante JR, et al.: Overall Survival and Durable Responses in Patients With BRAF V600-Mutant Metastatic Melanoma Receiving Dabrafenib Combined With Trametinib. J Clin Oncol. 2016; 34(8): 871–8. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nLarkin J, Ascierto PA, Dréno B, et al.: Combined vemurafenib and cobimetinib in BRAF-mutated melanoma. N Engl J Med. 2014; 371(20): 1867–76. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nAscierto PA, Schadendorf D, Berking C, et al.: MEK162 for patients with advanced melanoma harbouring NRAS or Val600 BRAF mutations: a non-randomised, open-label phase 2 study. Lancet Oncol. 2013; 14(3): 249–56. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nHandolias D, Hamilton AL, Salemi R, et al.: Clinical responses observed with imatinib or sorafenib in melanoma patients expressing mutations in KIT. Br J Cancer. 2010; 102(8): 1219–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHodi FS, Corless CL, Giobbie-Hurder A, et al.: Imatinib for melanomas harboring mutationally activated or amplified KIT arising on mucosal, acral, and chronically sun-damaged skin. J Clin Oncol. 2013; 31(26): 3182–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo J, Si L, Kong Y, et al.: Phase II, open-label, single-arm trial of imatinib mesylate in patients with metastatic melanoma harboring c-Kit mutation or amplification. J Clin Oncol. 2011; 29(21): 2904–9. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nKim KB, Eton O, Davis DW, et al.: Phase II trial of imatinib mesylate in patients with metastatic melanoma. Br J Cancer. 2008; 99(5): 734–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarvajal RD, Lawrence DP, Weber JS, et al.: Phase II Study of Nilotinib in Melanoma Harboring KIT Alterations Following Progression to Prior KIT Inhibition. Clin Cancer Res. 2015; 21(10): 2289–96. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPasseron T, Lacour J, Allegra M, et al.: Signalling and chemosensitivity assays in melanoma: is mutated status a prerequisite for targeted therapy? Exp Dermatol. 2011; 20(12): 1030–2. PubMed Abstract | Publisher Full Text\n\nMenzies AM, Yeh I, Botton T, et al.: Clinical activity of the MEK inhibitor trametinib in metastatic melanoma containing BRAF kinase fusion. Pigment Cell Melanoma Res. 2015; 28(5): 607–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw AT, Yeap BY, Solomon BJ, et al.: Effect of crizotinib on overall survival in patients with advanced non-small-cell lung cancer harbouring ALK gene rearrangement: a retrospective analysis. Lancet Oncol. 2011; 12(11): 1004–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw AT, Engleman JA: Ceritinib in ALK-rearranged non-small-cell lung cancer. N Engl J Med. 2014; 370(26): 2537–9. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCapper D, Preusser M, Habel A, et al.: Assessment of BRAF V600E mutation status by immunohistochemistry with a mutation-specific monoclonal antibody. Acta Neuropathol. 2011; 122(1): 11–9. PubMed Abstract | Publisher Full Text\n\nIlie M, Long-Mira E, Funck-Brentano E, et al.: Immunohistochemistry as a potential tool for routine detection of the NRAS Q61R mutation in patients with metastatic melanoma. J Am Acad Dermatol. 2015; 72(5): 786–93. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nAnderson S, Bloom KJ, Vallera DU, et al.: Multisite analytic performance studies of a real-time polymerase chain reaction assay for the detection of BRAF V600E mutations in formalin-fixed, paraffin-embedded tissue specimens of malignant melanoma. Arch Pathol Lab Med. 2012; 136(11): 1385–91. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14650",
"date": "28 Jun 2016",
"name": "Heinz Kutzner",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14651",
"date": "28 Jun 2016",
"name": "Lyn Duncan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14652",
"date": "28 Jun 2016",
"name": "Pedram Gerami",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1529
|
https://f1000research.com/articles/5-1524/v1
|
28 Jun 16
|
{
"type": "Review",
"title": "The regulation of hematopoietic stem cell populations",
"authors": [
"Hector Mayani"
],
"abstract": "Evidence presented over the last few years indicates that the hematopoietic stem cell (HSC) compartment comprises not just one but a number of different cell populations. Based on HSCs’ proliferation and engraftment potential, it has been suggested that there are two classes of HSC, with long- and short-term engraftment potential. HSC heterogeneity seems to involve differentiation capacities as well, since it has been shown that some HSC clones are able to give rise to both myeloid and lymphoid progeny, whereas others are lymphoid deficient. It has been recognized that HSC function depends on intrinsic cell regulators, which are modulated by external signals. Among the former, we can include transcription factors and non-coding RNAs as well as epigenetic modifiers. Among the latter, cytokines and extracellular matrix molecules have been implicated. Understanding the elements and mechanisms that regulate HSC populations is of significant relevance both in biological and in clinical terms, and research in this area still has to face several complex and exciting challenges.",
"keywords": [
"hematopoietic progenitor cells",
"precursor cells",
"hematopoietic stem cell niche"
],
"content": "Introduction\n\nThe term hematopoietic stem cell (HSC) refers to an immature cell, residing in the bone marrow, which is capable of both self-renewal and differentiation into all of the different blood cell types. Evidence presented over the last few years, however, indicates that the HSC compartment comprises not just one but a number of different cell populations. This, of course, has both biological and clinical implications. Accordingly, there is great interest in elucidating the identity of each of these cell populations and defining their biological differences and similarities in molecular, immunophenotypic, and functional terms.\n\n\nHSCs: basic principles\n\nAlthough the concept of a primitive, immature cell common to all of the different blood cell lineages (erythrocytes, leucocytes, and platelets) was presented in the first decades of the 20th century, it wasn’t until work by James Till and Ernest McCulloch in the early 1960s that the existence of such a stem cell was demonstrated1,2. The work by Till and McCulloch, in Toronto, Canada, together with that of Metcalf and colleagues, a few years later in Melbourne, Australia, showed that the hematopoietic system could be subdivided into four separate compartments: HSCs (comprising the most immature cells, those capable of self-renewal), hematopoietic progenitor cells (HPCs; those unable to self-renew, but with a large proliferative potential and multilineage, bilineage, or monolineage differentiation capacities), precursor cells (those immature cells that can already be identified through their morphology), and mature blood cells (those present in circulation)3.\n\nAlthough most of what we know about HSC biology comes from studies in animal models, mostly in mice, it has become evident that human HSCs follow similar biological patterns4,5. HSCs cannot be identified by morphological criteria; instead, their identification is based on both immunophenotypic analysis and functional assays6. Murine HSCs express antigens such as Sca-1, CD117, and CD150 and do not express CD48; human HSCs, on the other hand, express CD34, CD49f, CD90, and CD117 and do not express CD384,7. In both cases, HSCs do not express any lineage-restricted antigen, so they are referred to as lineage-negative (Lin-) cells4,7. It is noteworthy that within the human HSC pool a CD34-negative population has also been identified (CD34- CD38- Lin-), whose cycling status (dormancy) suggests that it is located at the apex of the HSC compartment8,9. Apart from the expression of specific cell surface markers, HSCs can also be identified by their ability to efflux certain fluorescent dyes, such as Rhodamine-123 (Rho) and Hoechst 33342; thus, they are known as Rho-/low cells or side population (SP) cells7. The latter form a characteristic cluster of events located off to the lower left side in dual wavelength fluorescence-activated cell sorting (FACS) dot-plot profiles.\n\nAssays to determine the number and functional integrity of HSCs include both in vivo and in vitro systems. The former consist of introducing HSCs into irradiated animals and determining the ability of such cells to repopulate the hematopoietic system of the host after several weeks post-transplant. This approach is based on the experiments described by Till and McCulloch1, although refined modifications have been introduced into the experimental system during the last few decades10. When using human HSCs, the recipient must be an immunodeficient animal (for instance, severe combined immunodeficient [SCID], non-obese diabetic [NOD]-SCID, or NOD SCID Gamma [NSG] mice), so there will not be rejection mediated by the immune cells of the host11. In vitro systems, on the other hand, are based on the ability of HSCs to initiate and sustain hematopoietic cell production for several weeks in cultures containing a stromal cell layer in the presence or in the absence of exogenous cytokines12,13. It is worth noting, however, that this latter method does not necessarily prove that the cells sustaining hematopoiesis in vitro are actual HSCs; thus, to date, the in vivo repopulation assay is the only method validated to detect and measure actual HSC function.\n\nIt is of particular importance to mention that recent work from Camargo’s and Rodewald’s groups has presented evidence indicating that during steady-state conditions, hematopoiesis is sustained by thousands of long-lived progenitors, rather than by actual HSCs14,15. These findings suggest that HSC function may not be as critical as previously thought for unperturbed hematopoiesis; in contrast, HSC activity seems to be of great relevance during stress hematopoiesis (e.g., post-hematopoietic cell transplants).\n\n\nOne or several HSC populations?\n\nSince its conception, the idea that the HSC compartment comprises a homogeneous cell population prevailed for some time; however, studies in animal (murine) models, reported over the last several years, demonstrated that some of the cells contained within the HSC pool are responsible for long-term engraftment, whereas others induce a transient, short-term engraftment4,6,16. Some groups have even suggested that, based on their engraftment potential, there are three classes of HSC: those with long-, intermediate, and short-term engraftment potential17. Thus, it is now recognized that the HSC population is heterogeneous, comprising several HSC subsets differing in their repopulation capacities and cycling properties18–20.\n\nAs expected, similar HSC populations also seem to exist in humans. Indeed, work from John Dick’s laboratory in Toronto has shown that human HSCs capable of long-term engraftment express CD34, CD49f, and CD90 and, of course, lack the expression of CD38 and any lineage-restricted antigen; thus, they are defined as CD34+ CD45RA- CD49f+ CD90+ CD38- Lin- cells (LT-HSCs). Loss of expression of CD49f and CD90 gives rise to transiently engrafting multipotent progenitors (MPPs [CD34+ CD45RA- CD49f- CD90- CD38- Lin- cells])21. Based on the genomic analysis performed by the same group on these populations, 70 genes were found to be differentially expressed between HSC subsets and MPPs, whereas 500–3000 genes were differentially expressed when comparing HSCs and more mature, committed progenitors21. All these findings clearly indicate the existence of not one but several populations within the HSC compartment.\n\nHSC heterogeneity seems to involve not only proliferative potentials, as mentioned above, but differentiation capacities as well. Work by different groups, including those of Muller-Sieburg and colleagues, Eaves and colleagues, and Suda and colleagues, has demonstrated that among murine HSCs, some clones are biased towards the production of myeloid cells (α-HSCs), whereas some are biased towards the production of lymphoid cells (γ/δ-HSCs); furthermore, others show a balanced capacity towards the production of both myeloid and lymphoid cells (β-HSCs)17,21–24. It is noteworthy that the relative proportion of each one of these HSC subsets varies throughout development. For instance, lymphoid-deficient cells (α-HSCs) are present at very low frequencies (<5% of all HSCs) in fetal liver, and their levels increase gradually with age, so that just before birth they constitute around 15% of all HSCs in fetal bone marrow. After birth, their levels correspond to 20% of HSCs, and in young adults their levels reach 25–30% of the HSC pool. In old mice, α-HSCs correspond to 45% of all HSCs25. The presence of α, β, γ, and δ HSCs in humans is less clear.\n\n\nHow are HSCs regulated?\n\nHSC viability, self-renewal, proliferation, commitment, and differentiation depend on both intrinsic and extrinsic elements. The former include a variety of regulatory molecules present within the cell, whereas the latter comprise different cell types and their products, which create the microenvironment in which HSCs grow. Thus, we can say that the function of HSCs is controlled by intrinsic cell regulators, which, in turn, are modulated by external signals26. Among the intrinsic regulators of stem cell function, we find nuclear transcription factors that control gene expression (for instance, the transcription factor SCL is essential for HSC survival, self-renewal, and quiescence27); molecular regulators of the cell cycle, including some cyclins and cyclin-dependent kinases28 (for instance, CDK6 is absent in long-term HSCs, which keeps them quiescent even in the presence of mitogenic stimulation; in contrast, short-term HSCs express high levels of CDK6, and this results in rapid entry into the cell cycle upon mitogenic stimulation29); the proteins responsible for setting up symmetric and asymmetric cell divisions, such as Musashi-230; molecules that act as mitotic clocks that set up the number of rounds of division (HSCs express high levels of telomerase, thus the length of their telomeres does not decrease as rapidly as in more mature cells31); and epigenetic regulators controlling the structure and organization of DNA and chromatin32.\n\nIn postnatal life, blood cell formation takes place primarily in the bone marrow. Here, stem cells are surrounded by different cell types, including stromal (e.g., mesenchymal stromal cells [MSCs], osteoblasts, fibroblasts, adipocytes, macrophages, and endothelial cells) and accessory (e.g., lymphocytes) cells. All of these different cell types form a unique environment, known as the HSC niche, that is responsible for providing HSCs with the right conditions for their growth33,34. Interestingly, recent evidence indicates that there are, in fact, several hematopoietic niches within the marrow microenvironment, including endosteal, vascular, and perivascular niches, which exert differential effects on HSCs35. The composition of each one of these niches is different, e.g., the endosteal niche consists mainly of osteoblasts, whereas the vascular niche consists of endothelial cells. The perivascular niche, in turn, contains both MSCs and Cxcl12-abundant reticular (CAR) cells35. Current evidence indicates that most of the HSCs residing in the bone marrow (around 85% of marrow HSCs) are located within 10 µm of a sinusoidal vessel36 and that cell fate is dictated mainly by elements of the perivascular niche35.\n\nThe cells that form part of the stem cell niche are able to produce and secrete a wide array of proteins – including extracellular matrix, cytokines, and chemokines – that influence stem cell behavior37. Cytokines exert their effects via specific molecules (receptors) located on the cell surface38, and they can be presented to their target cells as soluble or as membrane-bound molecules. The fact that some cytokines are presented as membrane-bound proteins implies that direct cell-to-cell interactions must take place between the cytokine-producing cell and the cytokine receptor-bearing cell. It has been suggested that the primary action of cytokines on stem cells is to prevent cell death and to promote cell division38.\n\n\nControlling HSC fate\n\nIn stem cell biology, the balance between self-renewal and differentiation is of key relevance. Cell fate decisions are associated with changes in gene expression and are controlled by the action of transcription factors39. Gene expression changes are usually accompanied (specifically preceded) by epigenetic changes in regulatory regions40. It is noteworthy, however, that the initial changes usually occur without de novo transcription and are mediated by the asymmetric distribution of cell fate determinants41.\n\nIn humans, several transcription factors have been associated with the HSC state, including ID genes, SOX8, SOX18, and NFIB. In contrast, factors such as MYC and IKZF1 have been implicated in differentiation into MPPs21. HOXB4 has also been found to be important in HSC biology, both in mice and in humans. Indeed, overexpression of such a factor in mouse HSCs induces symmetric divisions, which result in a 1000-fold expansion in HSC numbers42. BMI1, a polycomb-group factor, increases the multilineage potential of human HSCs, as well as their replating capacity; in contrast, bmi1 deletion results in loss of clonal potential43,44. Other genes whose expression favors self-renewal and confers increased repopulation potential include hes1 and hlf45, as well as notch46. Activation of certain genes and pathways has been implicated in the loss of HSC potential. For instance, activation of the mTOR pathway results in the loss of HSC self-renewal47; similarly, BATF activation decreases self-renewal capacity and induces lymphoid differentiation48.\n\nToday, we know that HSC fate choices are greatly influenced and controlled by epigenetic changes. For instance, an increase in H4K16Ac levels results in inhibition of Cdc42, and this, in turn, results in restoration of the B cell lineage output in aged HSCs49. Increased levels of H3K9me2 mark the onset of HSC lineage commitment, whereas inhibition of G9a improves HSC maintenance50. Non-coding RNAs are also key regulators of HSC biology. MiRNA-22, for example, is a powerful inducer of HSC maintenance and self-renewal51, and very recently miRNA-126 was shown to play a key role in the self-renewal capacity and outcome of HSCs52.\n\n\nImplications and challenges\n\nUnderstanding the elements and mechanisms that regulate HSC populations is of significant relevance at two different levels. On the one hand, it is important in our knowledge regarding blood cell production under both normal and pathological states; thus, it helps us decipher the steps and pathways that lead to disorders such as myelodysplasia or leukemia. Indeed, particular mutations in several transcription factors have been implicated in the pathophysiology of such disorders (reviewed in 4). On the other hand, it is important in the development of therapeutic strategies. For instance, knowledge of the Notch pathway has led to the development of laboratory strategies for the ex vivo expansion of HSCs and HPCs from human cord blood53,54.\n\nResearch on the regulation of HSC populations still has to face several challenges. One of the most obvious is related to the development of therapeutic strategies using specific regulatory molecules and intracellular pathways as targets; for example, WNT or Notch pathways. Another one would be trying to understand aging of the HSC pool at the single cell level. In this regard, it will be of the most importance in the development of single cell RNA technologies, some of which have already been worked out, to fully understand the processes of gene regulation in HSCs from young and old individuals. Promising and exciting years are yet to come.",
"appendix": "Competing interests\n\n\n\nThe author declares that he has no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nTill JE, McCulloch EA: A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. 1961. Radiat Res. 2012; 178(2): AV3–7. PubMed Abstract | Publisher Full Text\n\nTill JE, McCulloch EA: Hemopoietic stem cell differentiation. Biochim Biophys Acta. 1980; 605(4): 431–59. PubMed Abstract | Publisher Full Text\n\nMetcalf D: The regulatory control of hemopoietic populations. Prog Clin Biol Res. 1990; 356: 147–54. PubMed Abstract\n\nDoulatov S, Notta F, Laurenti E, et al.: Hematopoiesis: a human perspective. Cell Stem Cell. 2012; 10(2): 120–36. PubMed Abstract | Publisher Full Text\n\nSzilvassy SJ: The biology of hematopoietic stem cells. Arch Med Res. 2003; 34(6): 446–60. PubMed Abstract | Publisher Full Text\n\nEaves CJ: Hematopoietic stem cells: concepts, definitions, and the new reality. Blood. 2015; 125(17): 2605–13. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWognum AW, Eaves AC, Thomas TE: Identification and isolation of hematopoietic stem cells. Arch Med Res. 2003; 34(6): 461–75. PubMed Abstract | Publisher Full Text\n\nBhatia M, Bonnet D, Murdoch B, et al.: A newly discovered class of human hematopoietic cells with SCID-repopulating activity. Nat Med. 1998; 4(9): 1038–45. PubMed Abstract | Publisher Full Text\n\nAnjos-Afonso F, Currie E, Palmer HG, et al.: CD34- cells at the apex of the human hematopoietic stem cell hierarchy have distinctive cellular and molecular signatures. Cell Stem Cell. 2013; 13(2): 161–74. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMiller CL, Dykstra B, Eaves CJ: Characterization of mouse hematopoietic stem and progenitor cells. Curr Protoc Immunol. 2008; Chapter 22: Unit 22B.2. PubMed Abstract | Publisher Full Text\n\nKamel-Reid S, Dick JE: Engraftment of immune-deficient mice with human hematopoietic stem cells. Science. 1988; 242(4886): 1706–9. PubMed Abstract | Publisher Full Text\n\nSutherland HJ, Eaves CJ, Lansdorp PM, et al.: Differential regulation of primitive human hematopoietic cells in long-term cultures maintained on genetically engineered murine stromal cells. Blood. 1991; 78(3): 666–72. PubMed Abstract\n\nHao QL, Thiemann FT, Petersen D, et al.: Extended long-term culture reveals a highly quiescent and primitive human hematopoietic progenitor population. Blood. 1996; 88(9): 3306–13. PubMed Abstract\n\nSun J, Ramos A, Chapman B, et al.: Clonal dynamics of native haematopoiesis. Nature. 2014; 514(7522): 322–7. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBusch K, Klapproth K, Barile M, et al.: Fundamental properties of unperturbed haematopoiesis from stem cells in vivo. Nature. 2015; 518(7540): 542–6. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMiller PH, Knapp DJ, Eaves CJ: Heterogeneity in hematopoietic stem cell populations: implications for transplantation. Curr Opin Hematol. 2013; 20(4): 257–64. PubMed Abstract | Publisher Full Text\n\nEma H, Morita Y, Suda T: Heterogeneity and hierarchy of hematopoietic stem cells. Exp Hematol. 2014; 42(2): 74–82.e2. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFoudi A, Hochedlinger K, van Buren D, et al.: Analysis of histone 2B-GFP retention reveals slowly cycling hematopoietic stem cells. Nat Biotechnol. 2009; 27(1): 84–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQiu J, Papatsenko D, Niu X, et al.: Divisional history and hematopoietic stem cell function during homeostasis. Stem Cell Reports. 2014; 2(4): 473–90. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nTakizawa H, Regoes RR, Boddupalli CS, et al.: Dynamic variation in cycling of hematopoietic stem cells in steady state and inflammation. J Exp Med. 2011; 208(2): 273–84. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nNotta F, Doulatov S, Laurenti E, et al.: Isolation of single human hematopoietic stem cells capable of long-term multilineage engraftment. Science. 2011; 333(6039): 218–21. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMuller-Sieburg CE, Cho RH, Karlsson L, et al.: Myeloid-biased hematopoietic stem cells have extensive self-renewal capacity but generate diminished lymphoid progeny with impaired IL-7 responsiveness. Blood. 2004; 103(11): 4111–8. PubMed Abstract | Publisher Full Text\n\nMuller-Sieburg CE, Sieburg HB: The GOD of hematopoietic stem cells: a clonal diversity model of the stem cell compartment. Cell Cycle. 2006; 5(4): 394–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDykstra B, Kent D, Bowie M, et al.: Long-term propagation of distinct hematopoietic differentiation programs in vivo. Cell Stem Cell. 2007; 1(2): 218–29. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBabovic S, Eaves CJ: Hierarchical organization of fetal and adult hematopoietic stem cells. Exp Cell Res. 2014; 329(2): 185–91. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMayani H: A glance into somatic stem cell biology: basic principles, new concepts, and clinical relevance. Arch Med Res. 2003; 34(1): 3–15. PubMed Abstract | Publisher Full Text\n\nRojas-Sutterlin S, Lecuyer E, Hoang T: Kit and Scl regulation of hematopoietic stem cells. Curr Opin Hematol. 2014; 21(4): 256–64. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMatsumoto A, Nakayama KI: Role of key regulators of the cell cycle in maintenance of hematopoietic stem cells. Biochim Biophys Acta. 2013; 1830(2): 2335–44. PubMed Abstract | Publisher Full Text\n\nLaurenti E, Frelin C, Xie S, et al.: CDK6 levels regulate quiescence exit in human hematopoietic stem cells. Cell Stem Cell. 2015; 16(3): 302–13. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nPark SM, Deering RP, Lu Y, et al.: Musashi-2 controls cell fate, lineage bias, and TGF-β signaling in HSCs. J Exp Med. 2014; 211(1): 71–87. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nYui J, Chiu CP, Lansdorp PM: Telomerase activity in candidate stem cells from fetal liver and adult bone marrow. Blood. 1998; 91(9): 3255–62. PubMed Abstract\n\nMahmud N, Petro B, Baluchamy S, et al.: Differential effects of epigenetic modifiers on the expansion and maintenance of human cord blood stem/progenitor cells. Biol Blood Marrow Transplant. 2014; 20(4): 480–9. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nReagan MR, Rosen CJ: Navigating the bone marrow niche: translational insights and cancer-driven dysfunction. Nat Rev Rheumatol. 2016; 12(3): 154–68. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBoulais PE, Frenette PS: Making sense of hematopoietic stem cell niches. Blood. 2015; 125(17): 2621–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMorrison SJ, Scadden DT: The bone marrow niche for haematopoietic stem cells. Nature. 2014; 505(7483): 327–34. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nAcar M, Kocherlakota KS, Murphy MM, et al.: Deep imaging of bone marrow shows non-dividing stem cells are mainly perisinusoidal. Nature. 2015; 526(7571): 126–30. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMayani H, Guilbert LJ, Janowska-Wieczorek A: Biology of the hemopoietic microenvironment. Eur J Haematol. 1992; 49(5): 225–33. PubMed Abstract | Publisher Full Text\n\nMetcalf D: Hematopoietic cytokines. Blood. 2008; 111(2): 485–91. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nOrkin SH, Zon LI: Hematopoiesis: an evolving paradigm for stem cell biology. Cell. 2008; 132(4): 631–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBernstein BE, Mikkelsen TS, Xie X, et al.: A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell. 2006; 125(2): 315–26. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nNeumüller RA, Knoblich JA: Dividing cellular asymmetry: asymmetric cell division and its implications for stem cells and cancer. Genes Dev. 2009; 23(23): 2675–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAntonchuk J, Sauvageau G, Humphries RK: HOXB4-induced expansion of adult hematopoietic stem cells ex vivo. Cell. 2002; 109(1): 39–45. PubMed Abstract | Publisher Full Text\n\nRizo A, Dontje B, Vellenga E, et al.: Long-term maintenance of human hematopoietic stem/progenitor cells by expression of BMI1. Blood. 2008; 111(5): 2621–30. PubMed Abstract | Publisher Full Text\n\nRizo A, Olthof S, Han L, et al.: Repression of BMI1 in normal and leukemic human CD34+ cells impairs self-renewal and induces apoptosis. Blood. 2009; 114(8): 1498–505. PubMed Abstract | Publisher Full Text\n\nShojaei F, Trowbridge J, Gallacher L, et al.: Hierarchical and ontogenic positions serve to define the molecular basis of human hematopoietic stem cell behavior. Dev Cell. 2005; 8(5): 651–63. PubMed Abstract | Publisher Full Text\n\nVarnum-Finney B, Xu L, Brashem-Stein C, et al.: Pluripotent, cytokine-dependent, hematopoietic stem cells are immortalized by constitutive Notch1 signaling. Nat Med. 2000; 6(11): 1278–81. PubMed Abstract | Publisher Full Text\n\nHuang J, Nguyen-McCarty M, Hexner EO, et al.: Maintenance of hematopoietic stem cells through regulation of Wnt and mTOR pathways. Nat Med. 2012; 18(12): 1778–85. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWang J, Sun Q, Morita Y, et al.: A differentiation checkpoint limits hematopoietic stem cell self-renewal in response to DNA damage. Cell. 2012; 148(5): 1001–14. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFlorian MC, Dörr K, Niebel A, et al.: Cdc42 activity regulates hematopoietic stem cell aging and rejuvenation. Cell Stem Cell. 2012; 10(5): 520–30. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nChen X, Skutt-Kakaria K, Davison J, et al.: G9a/GLP-dependent histone H3K9me2 patterning during human hematopoietic stem cell lineage commitment. Genes Dev. 2012; 26(22): 2499–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong SJ, Ito K, Ala U, et al.: The oncogenic microRNA miR-22 targets the TET2 tumor suppressor to promote hematopoietic stem cell self-renewal and transformation. Cell Stem Cell. 2013; 13(1): 87–101. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nLechman ER, Gentner B, Ng SW, et al.: miR-126 Regulates Distinct Self-Renewal Outcomes in Normal and Malignant Hematopoietic Stem Cells. Cancer Cell. 2016; 29(2): 214–28. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nDelaney C, Heimfeld S, Brashem-Stein C, et al.: Notch-mediated expansion of human cord blood progenitor cells capable of rapid myeloid reconstitution. Nat Med. 2010; 16(2): 232–6. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMayani H: Notch signaling: from stem cell expansion to improving cord blood transplantation. Expert Rev Hematol. 2010; 3(4): 401–4. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14627",
"date": "28 Jun 2016",
"name": "Shannon McKinney-Freeman",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14626",
"date": "28 Jun 2016",
"name": "Dominique Bonnet",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1524
|
https://f1000research.com/articles/5-1514/v1
|
27 Jun 16
|
{
"type": "Review",
"title": "Management of postpartum haemorrhage",
"authors": [
"Marie Pierre Bonnet",
"Dan Benhamou",
"Marie Pierre Bonnet"
],
"abstract": "Postpartum Haemorrhage (PPH) is a major cause of maternal morbidity and mortality. Treatment of acquired coagulopathy observed in severe PPH is an important part of PPH management, but is mainly based on literature in trauma patients, and data thus should be interpreted with caution. This review describes recent advances in transfusion strategy and in the use of tranexamic acid and fibrinogen concentrates in women with PPH.",
"keywords": [
"coagulopathy",
"Postpartum Haemorrhage",
"transfusion strategy",
"tranexamic acid",
"fibrinogen"
],
"content": "Introduction\n\nPostpartum hemorrhage (PPH) is one of the most frequent life-threatening complications of going into labor and occurs mostly without any warning or predictive signs or symptoms and often in the absence of predisposing conditions. The main causes of PPH are uterine atony, retained placenta, and genital tract trauma. Abnormal placentation, placental abruption, and uterine rupture are less frequent but often responsible for severe PPH with acquired coagulopathy. PPH accounts for nearly one-quarter of all maternal deaths worldwide and an estimated 125,000 deaths occur each year1. Most of the time, these deaths due to obstetric hemorrhage are considered to be potentially preventable2,3. Maternal mortality is the end result of a worsening process, and PPH is also responsible for half of maternal morbidity4. The incidence of PPH has recently increased in most developed countries such as Canada, Australia, and the US and has been notably related to an increased use of oxytocin for labor augmentation and subsequent uterine atony5.\n\nCurrently, the therapeutic strategies for PPH management are largely standardized; in particular, obstetric, surgical, and radiological interventions play a life-saving role in PPH management. However, medical treatment, namely transfusion and a pro-hemostatic strategy, is also essential and has shown important changes in recent years. This review focuses on advances in transfusion strategy and on the use of pro-hemostatic agents such as tranexamic acid (TA) and fibrinogen concentrates in PPH.\n\n\nTransfusion strategy in postpartum hemorrhage\n\nOnly few data are available to guide transfusion management in the acute phase of PPH. The current guidelines are based mainly on the literature coming from trauma patients.\n\nIn trauma patients, several cohort studies have demonstrated a decrease in mortality associated with the administration of red blood cells (RBCs) and fresh frozen plasma (FFP) in a 1:1 ratio in the context of massive transfusion (10 units of RBCs or more)6–9. This strategy, which comes from a military setting and which is called damage control resuscitation, aims to administer coagulation factors as early as RBCs in order to treat blood loss but at the same time to prevent the coagulopathy observed in massively bleeding patients by limiting the use of crystalloids as volume replacement. These results are controversial. First, most of these studies were retrospective. Second, a survival bias cannot be ruled out10. Indeed, the more severe trauma patients died early and were not able to be timely transfused in FFP, inducing an increased mortality in patients receiving a transfusion with a low FFP-to-RBC ratio. The transfusion benefit of a high FFP-to-RBC ratio is not so clear in recent prospective cohort studies. In the prospective, observational, multicenter, major trauma transfusion (PROMMTT) study documenting the timing of transfusion during active resuscitation in 905 trauma patients, Holcomb et al. demonstrated that early and higher FFP-to-RBC ratios were associated with a decreased mortality in patients transfused with at least three units of RBCs during the first 24 hours after admission, but not at 30 days11. In a pragmatic randomized controlled trial from the same authors (the Pragmatic Randomized Optimal Platelet and Plasma Ratios [PROPPR] study), 680 severely injured patients requiring massive transfusion were randomly assigned between early administration of FFP, platelets, and RBCs in a 1:1:1 ratio compared with a 1:1:2 ratio. No significant difference in mortality at 24 hours or at 30 days was observed. However, in the 1:1:1 group, more patients achieved hemostasis and fewer experienced death due to exsanguination by 24 hours12. Finally, the use of FFP is associated with an increased incidence of complications such as post-injury multiple organ failure, acute respiratory distress, and infections, and the rate of complications increased with the quantities of FFP transfused13. For all of these reasons, the quality of the proofs in favor of a benefit in mortality with a transfusion in FFP and RBCs in a 1:1 ratio is considered low, and the optimal FFP-to-RBC ratio is still not determined. The recent European guidelines on coagulopathy management in trauma patients recommend an initial administration of plasma in patients with massive bleeding and, if further plasma is administered, an optimal FFP-to-RBC ratio of at least 1:2 is suggested. It is also recommended that FFP transfusion be avoided in trauma patients without substantial bleeding14.\n\nIn the obstetrical setting, there is no study on the impact of the FFP-to-RBC ratio on maternal morbidity and mortality. As severe PPH may result in secondary coagulopathy similar to what is described in massive blood loss in patients with trauma injury, some experts have proposed extending the transfusion guidelines described in trauma patients to women with severe PPH15. One study in obstetric hemorrhage evaluating this strategy of transfusion with a high FFP-to-RBC ratio was conducted by Alexander et al., who compared maternal outcomes between women who received whole blood only, women who received RBCs only, and women who received a combination of blood products16. In this study, complications attributable to hypovolemia were significantly increased in the combination group as compared with the whole blood and RBC groups. However, an indication bias could not be excluded, as hemorrhage was more severe in the group of women who received the combination of blood products, with larger quantities of RBCs transfused and an increased incidence of hysterectomy. For the moment, by analogy with transfusion strategy in trauma patients, transfusion with a high FFP-to-RBC ratio (between 1:2 and 1:1) should be dedicated only for women with PPH requiring massive transfusion.\n\nIn collaboration with the local blood bank, some teams have successfully developed the concept of massive transfusion protocol, which includes a transfusion package, early and repeated monitoring of hemostatic competence, and an intervention algorithm which is modeled on existing protocols used in the trauma service17,18. In case of massive PPH, a package consisting of six RBC units, at least four FFP units, and one apheresis platelet unit is immediately released (without waiting for laboratory results); this treatment is repeated as long as bleeding is not controlled.\n\nThe only strong recommendation on blood transfusion in PPH is that women receive RBCs as soon as possible in case of massive PPH. Because cross-matched blood is not always available, maternity units should have immediate access (within 5 minutes) to O-negative blood. If the need is less pressing, group-specific blood can be made available more quickly than fully cross-matched blood. Consequently, all maternity units should have their own reserve of blood products if there is no blood bank on-site19.\n\nFinally, it appears that, more than the predetermined ratio, the early treatment of coagulopathy with FFP and platelets determines maternal morbidity and mortality.\n\nUnfortunately, blood transfusion has its own adverse consequences. To decrease transfusion exposure and to control the bleeding, pro-hemostatic agents are used more and more often in women with PPH.\n\n\nPostpartum hemorrhage and tranexamic acid\n\nTA is an antifibrinolytic agent that inhibits the activation of plasminogen into plasmin. Its use is now clearly established for the control and prophylaxis of menorrhagia20. Its efficacy has also been proven in elective surgery such as orthopedic, vascular, hepatic, or urologic surgery and more recently in bleeding trauma patients21.\n\nA meta-analysis published in the British Medical Journal in 2012 pooled all of the randomized controlled trials comparing TA with no TA or placebo in surgical patients22. The results showed that TA reduced the probability of receiving a blood transfusion by one-third in elective surgery, but its effects on thromboembolic events and mortality remained uncertain.\n\nConcerning trauma patients, the CRASH-2 (Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage-2) multicenter trial showed that mortality was significantly decreased in trauma patients who received TA (1 g in 10 minutes, followed by a 1 g continuous infusion over the course of 8 hours) as compared with placebo. In particular, mortality due to hemorrhagic shock was decreased by 15%21.\n\nIn the obstetrical setting, the literature focuses mostly on the prophylactic use of TA to prevent PPH, particularly in the context of elective cesarean delivery. The most recent meta-analysis exploring the preventive effect on PPH and safety of TA versus placebo or no treatment was published in 2015 by the Cochrane database23. It included nine trials involving 2453 women who were at low risk of PPH and who were undergoing cesarean delivery and three trials with 832 women who delivered vaginally. These trials were of mixed quality. The results of the meta-analysis showed that overall the incidence of blood loss of greater than 500 mL was lower in women who received TA versus placebo or no intervention. Also, TA was effective in decreasing the incidence of blood loss of greater than 1000 mL in women who had undergone cesarean delivery but not vaginal birth. Mean blood loss within the first 2 hours postpartum was lower in women who received TA, and this effect was similar following vaginal and cesarean birth, but the mean difference in blood loss reduction was very moderate; the mean difference was 78 mL (95% confidence interval of 58 to 98 mL). Blood transfusion was less frequent in women receiving TA versus placebo or no intervention. Finally, the authors found that the use of TA was associated with only mild side effects, such as nausea, vomiting, and dizziness.\n\nDespite these encouraging results, the routine use of TA as a preventive treatment of PPH is currently not recommended and this was because of the moderate quality of the available evidence, the limited benefit on blood loss, and the lack of data on the thromboembolic risk.\n\nConcerning the use of TA as a curative treatment of PPH, only one randomized trial has been published until now. In this open-labeled controlled study, 144 women with a blood loss of 800 mL or more after vaginal delivery were randomly assigned to receive either 4 g of TA within 1 hour followed by 1 g per hour during 6 hours or not24. In the treatment group, total blood loss at 6 hours after PPH diagnosis was significantly lower, but the difference was questionable (170 versus 221 mL; P = 0.041). Duration of hemorrhage was also significantly shorter in the treatment group, and transfusion requirement and incidence of severe hemorrhage were lower.\n\nTo summarize, data on the effectiveness of TA in PPH are quite encouraging, but there is still only little reliable evidence coming from randomized controlled trials. However, owing to its low cost and low rate of side effects, the use of TA is currently recommended by several academic societies. For example, the most recently updated PPH treatment guidelines prepared by the World Health Organization state that TA (1 g over 5 minutes, repeated within 30 to 60 minutes if necessary) is recommended for the treatment of PPH if oxytocin and other uterotonics fail to stop bleeding or if it is thought that the bleeding may be partly due to trauma (weak recommendation)25. Consequently, there is an urgent need for clinical randomized trials of good quality before TA can be strongly recommended as a curative treatment in PPH.\n\nThe World Maternal Antifibrinolytic (WOMAN) trial conducted worldwide aims to determine the effect of the early administration of TA on mortality and hysterectomy rates as well as major complication rates in women with clinically diagnosed hemorrhage26. Inclusions are over now and the results of this study are due to be published later this year.\n\n\nFibrinogen concentrates in postpartum hemorrhage\n\nFibrinogen plays a critical role in achieving and maintaining hemostasis and is fundamental to effective clot formation. In the context of massive obstetric bleeding, fibrinogen is the first coagulation factor to decrease27; rapid fibrinolysis has also been described in some specific causes of PPH, such as placental abruption, placenta previa, genital tract trauma, and uterine atony.\n\nFibrinogen plasma level has been demonstrated to be a good predictor of PPH severity28,29. In the study by Charbit et al., a fibrinogen plasma level of 2 g/L or less had a 100% positive predictive value for severe PPH28. This study also demonstrated that the risk for severe PPH was 2.6 fold higher for each 1 g/L decrease in fibrinogen plasma level. Therefore, the assumption that fibrinogen supplementation could be beneficial to treat PPH has been made, although this is likely an over-interpretation of the study results. It should indeed be noted that the study by Charbit et al. was not randomized and did not demonstrate that decreased fibrinogen concentration was a causal factor of PPH severity. This study demonstrated only that decreased fibrinogen concentration was associated with PPH severity. Therefore, basing our hemostatic strategy on this argument requires further study.\n\nFFP transfusion is not the optimal agent for treating fibrinogen deficiency, as a volume of 30 mL/kg is necessary to increase the fibrinogen concentration by 1 g/L, inducing a high risk of fluid overload. In the past, fibrinogen therapy was usually given as cryoprecipitate, but owing to the potential viral contamination and variable concentration of fibrinogen in cryoprecipitate, human plasma-derived fibrinogen concentrates are now available in most countries but not everywhere. For example, in the UK, the only licensed source of fibrinogen is FFP or cryoprecipitate, which also contains von Willebrand factor, factor VIII, factor XIII, and fibronectin. Fibrinogen concentrates offer rapid restoration of the fibrinogen concentration with a small-volume infusion, for a comparable cost, and with a minimal preparation time. Fibrinogen concentrates are considered by many to be preferable to cryoprecipitate, although there are no studies comparing the efficacy of these two products30.\n\nThe efficacy and safety of fibrinogen concentrates have been proven in congenital fibrinogen deficiencies31. Additionally, some in vitro and animal studies with thromboelastography monitoring have shown that the addition of fibrinogen concentrates corrects the coagulation disorders induced by experimental hemodilution32,33. Clinical data on the efficacy of fibrinogen concentrates in the management of hemorrhage are still scarce. Overall, fewer than 10 randomized controlled trials have explored the potential benefit of fibrinogen concentrates in terms of transfusion requirement and correction of hemostasis disorders analyzed either by traditional hemostatic monitoring or by point-of-care (POC) viscoelastic methods34–41. These trials included 384 patients overall and were all performed in the setting of perioperative bleeding in scheduled cardiovascular surgery, except one trial including patients undergoing radical cystectomy35. Overall, five trials found a decrease in transfusion requirement in patients who received fibrinogen concentrates34,35,39–41, and five trials described an increased clot firmness measured by thromboelastometry35,38–41. However, only one study found a significant reduction in blood loss, which was not associated with a transfusion-sparing effect37, and none of these trials found a significant difference in mortality rate. These trials have several methodological flaws: small sample size, no prolonged follow-up, no intention-to-treat analysis, and no or poor blindness design. The protocol of fibrinogen concentrate administration differed between studies in terms of dose and of therapeutic target (prophylactic or curative treatment) as well as the control group. Consequently, it is difficult to extend these results to the obstetrical setting.\n\nIn trauma patients, only observational studies have been published42. In retrospective studies, the administration of fibrinogen concentrates was associated with a decrease in transfusion needs and with the correction of biologic hemostatic disorders43–46. In one of these studies, the observed mortality was lower than predicted mortality for patients who received fibrinogen concentrates47. However, these retrospective studies also have several methodological flaws. In particular, the severity of hemorrhage and confounding factors concerning blood loss volume were not taken into account, inducing an indication bias. No randomized controlled trial on the impact of fibrinogen concentrates in trauma patients has yet been published. To summarize, data on fibrinogen concentrate efficacy and safety in bleeding trauma patients are too limited for a conclusion to be drawn.\n\nThe use of fibrinogen concentrate in PPH has been explored in seven observational studies, in which a total of 222 women participated43,48–51. In six studies, a significant increase in fibrinogen plasma level was described after the administration of fibrinogen concentrate, but without a control group, the efficacy of fibrinogen concentrates cannot be determined. The only controlled study is a before-and-after study of 77 women with PPH52. In this retrospective study, maternal outcomes were compared between women who received cryoprecipitate (n = 14) and those who received fibrinogen concentrate (n = 20). The authors did not find any difference in blood loss, transfusion requirement, or need for a surgical hemostatic procedure. The first randomized controlled trial investigating the use of fibrinogen concentrate in PPH, in which 229 women participated, was published this year53. The “FIB-PPH Trial”, as it is known, is a Danish multicenter placebo-controlled, double-blinded clinical trial evaluating whether initial treatment with fibrinogen concentrate (2 g) reduces the need for allogenic blood transfusion in PPH. No difference was observed between the two groups in RBC transfusion requirement up to 6 weeks postpartum or in any of the secondary predefined outcomes (total blood loss, total amount of blood transfused, occurrence or rebleeding, low hemoglobin level, a composite outcome of severe PPH, and RBC transfusion within 4 hours, 24 hours, and 7 days). However, women included in this trial had no acquired hypofibrinogenemia. Consequently, this study could draw conclusions only on the inefficacy of the use of fibrinogen concentrates as a pre-emptive treatment for severe PPH in patients with normofibrinogenemia.\n\nFinally, the literature on the use of fibrinogen concentrate in non-obstetric hemorrhage only moderately suggests an efficacy on transfusion requirement and morbidity without formally proving it. Even if it appears to be a promising therapeutic, there is still no strong evidence that the use of fibrinogen concentrate would improve maternal outcomes in severe PPH. Moreover, the risk of thromboembolic events associated with the use of fibrinogen concentrate has never been explored in this context. Therefore, we still need valid data before administration of fibrinogen concentrate as a curative treatment of PPH can be firmly recommended.\n\n\nAdditional strategies\n\nRecombinant human FVIIa generated great hope several years ago when early case reports suggested immediate efficacy in refractory PPH54. Unfortunately, randomized trials in trauma and more recently in obstetrics have shown only a moderate decrease in blood product consumption but no survival benefit, whereas the risk of thrombotic events seems to increase significantly55.\n\nProthrombin complex concentrates contain several important coagulation factors and it has been suggested that they could replace FFP. This has been shown mainly in case reports or series in which coagulation factor deficit was detected by using POC viscoelastic tests in trauma47 or traditional hemostatic tests in obstetric patients56.\n\n\nViscoelastic point-of-care coagulation monitoring\n\nIn current strategies, drugs and blood products are administered very early and with aggressive protocols. These are mostly “blind” techniques, since drugs or blood products are given either prophylactically or in response to severe blood loss but very often before the return of laboratory results can inform the physician of coagulation abnormalities. Moreover, blood product and, especially, FFP administration is aimed at globally correcting coagulation without a precise target.\n\nHowever, it has been shown in recent years that it is possible to obtain coagulation tests results very early by using POC devices which are based mainly on viscoelastic techniques: thromboelastography (TEG) or thromboelastometry (ROTEM). These biological techniques can provide results within minutes and precisely inform the physician of the main hemostatic abnormalities. Physicians can thus direct treatment against precise targets and avoid (or reduce) the use of blood products. As shown above, hyperfibrinolysis and decreased fibrinogen concentration are important coagulation defects in massive hemorrhage and especially in obstetrics27 and can be detected early by using viscoelastic techniques57. Thus, early administration of fibrinogen concentrates can be guided by POC results rather than by the theoretical premise that these defects are very frequent. In a before-and-after study, Mallaiah et al. recently showed a significant decrease in blood product component use and a reduced incidence of circulatory overload with ROTEM-guided fibrinogen concentrate administration in major obstetric hemorrhage as compared with traditional care58. There is currently no randomized study in obstetrics showing that this strategy decreases blood product use and improves outcomes, but small randomized studies performed in cardiovascular surgery have described positive results39,41. Increased use of these viscoelastic tests is likely to occur soon because although these tests remain relatively costly and there is a need for quality control of these machines, new versions, which are as easy to use as POC hemoglobin measurement tests, have been recently released.\n\n\nTeamwork and safety issues\n\nPPH is a typical obstetric emergency situation that can develop rapidly and unexpectedly. Health-care professionals taking care of obstetric emergencies act as a team in which each provider uses his or her specific competencies. Moreover, because each obstetric emergency situation is a relatively rare event, even in high-level reference centers, providers have relatively few opportunities to train by self-experience and to evaluate and discuss how previous cases have been managed. During any emergency situation, communication and organizing the process of care are difficult tasks. It has been recognized that in many cases there is no clear leadership59, and poor teamwork has been recognized as a major cause of poor outcome60. For each team member, non-technical skills thus represent an important component of competency. Flin and Maran have described non-technical skills as two cognitive competencies (situational awareness and decision making) and two social competencies (teamwork and leadership)61. Most of these competencies are universal and should be practiced by every health-care provider whatever his or her profession or grade. Unfortunately, using these four skills is difficult and does not come naturally to most humans. Recognition of these deficiencies and subsequent training are thus essential. Unfortunately, traditional teaching is almost ineffective to improve patients’ outcomes. Recently developed strategies have emerged to facilitate adoption of these non-technical skills in clinical practice. Interprofessional education is a relatively recent concept and is said to occur ‘when two or more professions learn with, from and about each other to improve collaboration and the quality of care’62. Interprofessional education is believed to be important for undergraduate students63 but also for professionals working in clinical units to develop or maintain interprofessional collaboration. Team training is gaining popularity, as it is now recognized to improve quality of care64. Team training as well as interprofessional education for students can be done through formal courses and meetings, but many studies have shown that simulation is effective to improve communication, teamwork, and patients’ outcomes65,66. Quality of care can also be improved by writing protocols that are made available to all providers67 and through retrospective audits, which reduce the incidence of severe PPH and improve the application of recommendations68.\n\n\nConclusions\n\nUntil now, no study has proven that a specific transfusion strategy or the use of any pro-hemostatic agent would improve maternal outcomes in the context of PPH. Levels of evidence of TA and fibrinogen concentrate efficacy and safety in PPH are low. Randomized controlled trials in the context of severe PPH are difficult to perform, but there is room for studies of good quality to explore these therapeutic options.\n\n\nAbbreviations\n\nFFP, fresh frozen plasma; POC, point-of-care; PPH, postpartum hemorrhage; RBC, red blood cell; TA, tranexamic acid.",
"appendix": "Competing interests\n\n\n\nMarie-Pierre Bonnet declares that she has no competing interests. Dan Benhamou has acted as a consultant for Octapharma (Lachen, Switzerland).\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHogan MC, Foreman KJ, Naghavi M, et al.: Maternal mortality for 181 countries, 1980–2008: a systematic analysis of progress towards Millennium Development Goal 5. Lancet. 2010; 375(9726): 1609–23. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nKnight M, Kenyon S, Brocklehurst P, et al.: Saving Lives, Improving Mothers' Care - Lessons learned to inform future maternity care from the UK and Ireland Confidential Enquiries into Maternal Deaths and Morbidity 2009–12. Oxford: University of Oxford, 2014. Reference Source\n\nEnquète Nationale Confidentielle sur les Morts Maternelles, France 2007–2009.Rapport du Comité National d'experts sur la Mortalité Maternelle (CNEMM). 2013. Reference Source\n\nZhang W, Alexander S, Bouvier-Colle M, et al.: Incidence of severe pre-eclampsia, postpartum haemorrhage and sepsis as a surrogate marker for severe maternal morbidity in a European population-based study: the MOMS-B survey. BJOG. 2005; 112(1): 89–96. PubMed Abstract | Publisher Full Text\n\nBelghiti J, Kayem G, Dupont C, et al.: Oxytocin during labour and risk of severe postpartum haemorrhage: a population-based, cohort-nested case-control study. BMJ Open. 2011; 1(2): e000514. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBorgman MA, Spinella PC, Perkins JG, et al.: The ratio of blood products transfused affects mortality in patients receiving massive transfusions at a combat support hospital. J Trauma. 2007; 63(4): 805–13. PubMed Abstract | Publisher Full Text\n\nHolcomb JB, Jenkins D, Rhee P, et al.: Damage control resuscitation: directly addressing the early coagulopathy of trauma. J Trauma. 2007; 62(2): 307–10. PubMed Abstract | Publisher Full Text\n\nMaegele M, Lefering R, Paffrath T, et al.: Changes in transfusion practice in multiple injury between 1993 and 2006: a retrospective analysis on 5389 patients from the German Trauma Registry. Transfus Med. 2009; 19(3): 117–24. PubMed Abstract | Publisher Full Text\n\nBorgman MA, Spinella PC, Holcomb JB, et al.: The effect of FFP:RBC ratio on morbidity and mortality in trauma patients based on transfusion prediction score. Vox Sang. 2011; 101(1): 44–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHo AM, Dion PW, Yeung JH, et al.: Prevalence of survivor bias in observational studies on fresh frozen plasma: erythrocyte ratios in trauma requiring massive transfusion. Anesthesiology. 2012; 116(3): 716–28. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nHolcomb JB, del Junco DJ, Fox EE, et al.: The prospective, observational, multicenter, major trauma transfusion (PROMMTT) study: comparative effectiveness of a time-varying treatment with competing risks. JAMA Surg. 2013; 148(2): 127–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHolcomb JB, Tilley BC, Baraniuk S, et al.: Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial. JAMA. 2015; 313(5): 471–82. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWatson GA, Sperry JL, Rosengart MR, et al.: Fresh frozen plasma is independently associated with a higher risk of multiple organ failure and acute respiratory distress syndrome. J Trauma. 2009; 67(2): 221–7; discussion 228–30. PubMed Abstract | Publisher Full Text\n\nSpahn DR, Bouillon B, Cerny V, et al.: Management of bleeding and coagulopathy following major trauma: an updated European guideline. Crit Care. 2013; 17(2): R76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaule I, Hawkins N: Transfusion practice in major obstetric haemorrhage: lessons from trauma. Int J Obstet Anesth. 2012; 21(1): 79–83. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nAlexander JM, Sarode R, McIntire DD, et al.: Whole blood in the management of hypovolemia due to obstetric hemorrhage. Obstet Gynecol. 2009; 113(6): 1320–6. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBurtelow M, Riley E, Druzin M, et al.: How we treat: management of life-threatening primary postpartum hemorrhage with a standardized massive transfusion protocol. Transfusion. 2007; 47(9): 1564–72. PubMed Abstract | Publisher Full Text\n\nSkupski DW, Lowenwirt IP, Weinbaum FI, et al.: Improving hospital systems for the care of women with major obstetric hemorrhage. Obstet Gynecol. 2006; 107(5): 977–83. PubMed Abstract | Publisher Full Text\n\nButwick AJ, Goodnough LT: Transfusion and coagulation management in major obstetric hemorrhage. Curr Opin Anaesthesiol. 2015; 28(3): 275–84. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMatteson KA, Rahn DD, Wheeler TL, et al.: Nonsurgical management of heavy menstrual bleeding: a systematic review. Obstet Gynecol. 2013; 121(3): 632–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShakur H, Roberts I, Bautista R, et al.: Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet. 2010; 376(9734): 23–32. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nKer K, Edwards P, Perel P, et al.: Effect of tranexamic acid on surgical bleeding: systematic review and cumulative meta-analysis. BMJ. 2012; 344: e3054. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNovikova N, Hofmeyr GJ, Cluver C: Tranexamic acid for preventing postpartum haemorrhage. Cochrane Database Syst Rev. 2015; (6): CD007872. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nDucloy-Bouthors AS, Jude B, Duhamel A, et al.: High-dose tranexamic acid reduces blood loss in postpartum haemorrhage. Crit Care. 2011; 15(2): R117. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWHO Guidelines Approved by the Guidelines Review Committee: WHO recommendations for the prevention and treatment of postpartum haemorrhage. Geneva, Switzerland, 2012. PubMed Abstract\n\nShakur H, Elbourne D, Gülmezoglu M, et al.: The WOMAN Trial (World Maternal Antifibrinolytic Trial): tranexamic acid for the treatment of postpartum haemorrhage: an international randomised, double blind placebo controlled trial. Trials. 2010; 11: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Lloyd L, Bovington R, Kaye A, et al.: Standard haemostatic tests following major obstetric haemorrhage. Int J Obstet Anesth. 2011; 20(2): 135–41. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCharbit B, Mandelbrot L, Samain E, et al.: The decrease of fibrinogen is an early predictor of the severity of postpartum hemorrhage. J Thromb Haemost. 2007; 5(2): 266–73. PubMed Abstract | Publisher Full Text\n\nCortet M, Deneux-Tharaux C, Dupont C, et al.: Association between fibrinogen level and severity of postpartum haemorrhage: secondary analysis of a prospective trial. Br J Anaesth. 2012; 108(6): 984–9. PubMed Abstract | Publisher Full Text\n\nCollis RE, Collins PW: Haemostatic management of obstetric haemorrhage. Anaesthesia. 2015; 70(Suppl 1): 78–86, e27–8. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBornikova L, Peyvandi F, Allen G, et al.: Fibrinogen replacement therapy for congenital fibrinogen deficiency. J Thromb Haemost. 2011; 9(9): 1687–704. PubMed Abstract | Publisher Full Text\n\nFenger-Eriksen C, Anker-Møller E, Heslop J, et al.: Thrombelastographic whole blood clot formation after ex vivo addition of plasma substitutes: improvements of the induced coagulopathy with fibrinogen concentrate. Br J Anaesth. 2005; 94(3): 324–9. PubMed Abstract | Publisher Full Text\n\nFries D, Krismer A, Klingler A, et al.: Effect of fibrinogen on reversal of dilutional coagulopathy: a porcine model. Br J Anaesth. 2005; 95(2): 172–7. PubMed Abstract | Publisher Full Text\n\nCui Y, Hei F, Long C, et al.: Perioperative monitoring of thromboelastograph on blood protection and recovery for severely cyanotic patients undergoing complex cardiac surgery. Artif Organs. 2010; 34(11): 955–60. PubMed Abstract | Publisher Full Text\n\nFenger-Eriksen C, Jensen TM, Kristensen BS, et al.: Fibrinogen substitution improves whole blood clot firmness after dilution with hydroxyethyl starch in bleeding patients undergoing radical cystectomy: a randomized, placebo-controlled clinical trial. J Thromb Haemost. 2009; 7(5): 795–802. PubMed Abstract | Publisher Full Text\n\nGalas FR, de Almeida JP, Fukushima JT, et al.: Hemostatic effects of fibrinogen concentrate compared with cryoprecipitate in children after cardiac surgery: a randomized pilot trial. J Thorac Cardiovasc Surg. 2014; 148(4): 1647–55. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nKarlsson M, Ternström L, Hyllner M, et al.: Prophylactic fibrinogen infusion reduces bleeding after coronary artery bypass surgery. A prospective randomised pilot study. Thromb Haemost. 2009; 102(1): 137–44. PubMed Abstract | Publisher Full Text\n\nLancé MD, Ninivaggi M, Schols SE, et al.: Perioperative dilutional coagulopathy treated with fresh frozen plasma and fibrinogen concentrate: a prospective randomized intervention trial. Vox Sang. 2012; 103(1): 25–34. PubMed Abstract | Publisher Full Text\n\nRahe-Meyer N, Solomon C, Hanke A, et al.: Effects of fibrinogen concentrate as first-line therapy during major aortic replacement surgery: a randomized, placebo-controlled trial. Anesthesiology. 2013; 118(1): 40–50. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nTanaka KA, Egan K, Szlam F, et al.: Transfusion and hematologic variables after fibrinogen or platelet transfusion in valve replacement surgery: preliminary data of purified lyophilized human fibrinogen concentrate versus conventional transfusion. Transfusion. 2014; 54(1): 109–18. PubMed Abstract | Publisher Full Text\n\nRanucci M, Baryshnikova E, Crapelli GB, et al.: Randomized, double-blinded, placebo-controlled trial of fibrinogen concentrate supplementation after complex cardiac surgery. J Am Heart Assoc. 2015; 4(6): e002066. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nAubron C, Reade MC, Fraser JF, et al.: Efficacy and safety of fibrinogen concentrate in trauma patients--a systematic review. J Crit Care. 2014; 29(3): 471.e11–7. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFenger-Eriksen C, Lindberg-Larsen M, Christensen AQ, et al.: Fibrinogen concentrate substitution therapy in patients with massive haemorrhage and low plasma fibrinogen concentrations. Br J Anaesth. 2008; 101(6): 769–73. PubMed Abstract | Publisher Full Text\n\nInnerhofer P, Westermann I, Tauber H, et al.: The exclusive use of coagulation factor concentrates enables reversal of coagulopathy and decreases transfusion rates in patients with major blunt trauma. Injury. 2013; 44(2): 209–16. PubMed Abstract | Publisher Full Text\n\nNienaber U, Innerhofer P, Westermann I, et al.: The impact of fresh frozen plasma vs coagulation factor concentrates on morbidity and mortality in trauma-associated haemorrhage and massive transfusion. Injury. 2011; 42(7): 697–701. PubMed Abstract | Publisher Full Text\n\nSchöchl H, Nienaber U, Maegele M, et al.: Transfusion in trauma: thromboelastometry-guided coagulation factor concentrate-based therapy versus standard fresh frozen plasma-based therapy. Crit Care. 2011; 15(2): R83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchöchl H, Nienaber U, Hofer G, et al.: Goal-directed coagulation management of major trauma patients using thromboelastometry (ROTEM)-guided administration of fibrinogen concentrate and prothrombin complex concentrate. Crit Care. 2010; 14(2): R55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuasch E, Alsina E, Díez J, et al.: Hemorragia obstétrica: estudio observacional sobre 21.726 partos en 28 meses. Rev Esp Anestesiol Reanim. 2009; 56(3): 139–46. Publisher Full Text\n\nWeinkove R, Rangarajan S: Fibrinogen concentrate for acquired hypofibrinogenaemic states. Transfus Med. 2008; 18(3): 151–7. PubMed Abstract | Publisher Full Text\n\nWeiss G, Lison S, Glaser M, et al.: Observational study of fibrinogen concentrate in massive hemorrhage: evaluation of a multicenter register. Blood Coagul Fibrinolysis. 2011; 22(8): 727–34. PubMed Abstract | Publisher Full Text\n\nKikuchi M, Itakura A, Miki A, et al.: Fibrinogen concentrate substitution therapy for obstetric hemorrhage complicated by coagulopathy. J Obstet Gynaecol Res. 2013; 39(4): 770–6. PubMed Abstract | Publisher Full Text\n\nAhmed S, Harrity C, Johnson S, et al.: The efficacy of fibrinogen concentrate compared with cryoprecipitate in major obstetric haemorrhage--an observational study. Transfus Med. 2012; 22(5): 344–9. PubMed Abstract | Publisher Full Text\n\nWikkelsø AJ, Edwards HM, Afshari A, et al.: Pre-emptive treatment with fibrinogen concentrate for postpartum haemorrhage: randomized controlled trial. Br J Anaesth. 2015; 114(4): 623–33. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nBouwmeester FW, Jonkhoff AR, Verheijen RH, et al.: Successful treatment of life-threatening postpartum hemorrhage with recombinant activated factor VII. Obstet Gynecol. 2003; 101(6): 1174–6. PubMed Abstract | Publisher Full Text\n\nLavigne-Lissalde G, Aya AG, Mercier FJ, et al.: Recombinant human FVIIa for reducing the need for invasive second-line therapies in severe refractory postpartum hemorrhage: a multicenter, randomized, open controlled trial. J Thromb Haemost. 2015; 13(4): 520–9. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nGlynn JC, Plaat F: Prothrombin complex for massive obstetric haemorrhage. Anaesthesia. 2007; 62(2): 202–3. PubMed Abstract | Publisher Full Text\n\nHuissoud C, Carrabin N, Audibert F, et al.: Bedside assessment of fibrinogen level in postpartum haemorrhage by thrombelastometry. BJOG. 2009; 116(8): 1097–102. PubMed Abstract | Publisher Full Text\n\nMallaiah S, Barclay P, Harrod I, et al.: Introduction of an algorithm for ROTEM-guided fibrinogen concentrate administration in major obstetric haemorrhage. Anaesthesia. 2015; 70(2): 166–75. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nGuise JM, Segel SY, Larison K, et al.: STORC safety initiative: a multicentre survey on preparedness & confidence in obstetric emergencies. Qual Saf Health Care. 2010; 19(6): e41. PubMed Abstract | Publisher Full Text\n\nMazzocco K, Petitti DB, Fong KT, et al.: Surgical team behaviors and patient outcomes. Am J Surg. 2009; 197(5): 678–85. PubMed Abstract | Publisher Full Text\n\nFlin R, Maran N: Basic concepts for crew resource management and non-technical skills. Best Pract Res Clin Anaesthesiol. 2015; 29(1): 27–39. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\n(CAIPE) CftAoIE. 2016. Reference Source\n\nLapkin S, Levett-Jones T, Gilligan C: A systematic review of the effectiveness of interprofessional education in health professional programs. Nurse Educ Today. 2013; 33(2): 90–102. PubMed Abstract | Publisher Full Text\n\nWeaver SJ, Dy SM, Rosen MA: Team-training in healthcare: a narrative synthesis of the literature. BMJ Qual Saf. 2014; 23(5): 359–72. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nFung L, Boet S, Bould MD, et al.: Impact of crisis resource management simulation-based training for interprofessional and interdisciplinary teams: A systematic review. J Interprof Care. 2015; 29(5): 433–44. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPhipps MG, Lindquist DG, McConaughey E, et al.: Outcomes from a labor and delivery team training program with simulation component. Am J Obstet Gynecol. 2012; 206(1): 3–9. PubMed Abstract | Publisher Full Text\n\nShields LE, Smalarz K, Reffigee L, et al.: Comprehensive maternal hemorrhage protocols improve patient safety and reduce utilization of blood products. Am J Obstet Gynecol. 2011; 205(4): 368.e1–8. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nDupont C, Deneux-Tharaux C, Touzet S, et al.: Clinical audit: a useful tool for reducing severe postpartum haemorrhages? Int J Qual Health Care. 2011; 23(5): 583–9. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "14611",
"date": "27 Jun 2016",
"name": "Brendan Carvalho",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14612",
"date": "27 Jun 2016",
"name": "Felicity Plaat",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1514
|
https://f1000research.com/articles/5-1513/v1
|
27 Jun 16
|
{
"type": "Review",
"title": "Pathogenesis of myasthenia gravis: update on disease types, models, and mechanisms",
"authors": [
"William D. Phillips",
"Angela Vincent",
"William D. Phillips"
],
"abstract": "Myasthenia gravis is an autoimmune disease of the neuromuscular junction (NMJ) caused by antibodies that attack components of the postsynaptic membrane, impair neuromuscular transmission, and lead to weakness and fatigue of skeletal muscle. This can be generalised or localised to certain muscle groups, and involvement of the bulbar and respiratory muscles can be life threatening. The pathogenesis of myasthenia gravis depends upon the target and isotype of the autoantibodies. Most cases are caused by immunoglobulin (Ig)G1 and IgG3 antibodies to the acetylcholine receptor (AChR). They produce complement-mediated damage and increase the rate of AChR turnover, both mechanisms causing loss of AChR from the postsynaptic membrane. The thymus gland is involved in many patients, and there are experimental and genetic approaches to understand the failure of immune tolerance to the AChR. In a proportion of those patients without AChR antibodies, antibodies to muscle-specific kinase (MuSK), or related proteins such as agrin and low-density lipoprotein receptor-related protein 4 (LRP4), are present. MuSK antibodies are predominantly IgG4 and cause disassembly of the neuromuscular junction by disrupting the physiological function of MuSK in synapse maintenance and adaptation. Here we discuss how knowledge of neuromuscular junction structure and function has fed into understanding the mechanisms of AChR and MuSK antibodies. Myasthenia gravis remains a paradigm for autoantibody-mediated conditions and these observations show how much there is still to learn about synaptic function and pathological mechanisms.",
"keywords": [
"Myasthenia gravis",
"neuromuscular junction",
"immunoglobulin",
"AChR"
],
"content": "Introduction\n\nMyasthenia gravis (MG) is a paradigm autoantibody-mediated disease. Antibodies to the acetylcholine receptor (AChR) are found in 85% of patients with generalised muscle weakness and in 50% of those with purely ocular involvement1. There is ample evidence from in vitro and in vivo approaches that these antibodies are pathogenic. AChR antibodies are typically of the immunoglobulin (Ig)G1 and IgG3 (human) subclasses, can lead to complement-mediated attack, and, being able to bind divalently to adjacent AChRs on the muscle surface, can also increase the rate of AChR internalisation (for a review of the earlier history of MG research, see 2). The resulting loss of AChRs at the neuromuscular junction (NMJ) impairs neuromuscular transmission (see Figure 1). This becomes clinically evident as fatigue and muscle weakness. In a minority of patients, however, the autoantibodies instead bind to muscle-specific kinase (MuSK). MuSK is a transmembrane tyrosine receptor kinase that is crucial for the development and maintenance of AChR clusters at the NMJ. These antibodies are clearly pathogenic, but the mechanisms are only recently beginning to be unravelled3.\n\n(A) Healthy neuromuscular transmission. The nerve terminal can release the contents of each vesicle (quanta) of acetylcholine by exocytosis. Spontaneous release of single quanta of acetylcholine activates the intrinsic cation channels of acetylcholine receptors (AChRs) in the postsynaptic membrane to produce a small, transient depolarisation called a miniature endplate potential (mEPP). The nerve action potential opens voltage-gated calcium channels (VGCCs) and triggers exocytosis of many quanta of acetylcholine, simultaneously producing the (much larger) EPP. In healthy individuals, the amplitude of the EPP is more than enough to reach the threshold required to activate the postsynaptic voltage-gated sodium channels (VGNaCs) and generate a muscle action potential. (B) The myasthenia gravis neuromuscular junction. AChR antibodies (mainly immunoglobulin [Ig]G1) activate complement, resulting in membrane attack complex-mediated damage to the post-junctional membrane architecture. The postsynaptic AChR numbers are depleted by divalent antibodies inducing AChR internalisation. The loss of AChRs results in smaller mEPP and EPP amplitudes. The EPP may not reach threshold, especially when the nerve is repetitively activated. Abbreviations: AChE, acetylcholinesterase\n\nThe pathogenic actions of autoantibodies at the level of the NMJ can be studied by a variety of techniques. Experiments on cultured muscle-like cells (TE671, C2C12 myotubes; outlined in 4) help define post-synaptic mechanisms in both AChR and MuSK antibody forms of the disease, but in vivo models are required to study the effects of the antibodies on the electrophysiology of neuromuscular transmission. A microelectrode can be used to record the membrane electrical potential of the muscle fibre near the NMJ. When the nerve is electrically stimulated, neuromuscular transmission can be detected as a brief rise in membrane potential, called the endplate potential (EPP5). Spontaneous miniature EPPs (mEPPs), which are much smaller in amplitude than the (evoked) EPP, provide a measure of the response of the postsynaptic AChRs to release of a single synaptic vesicle-load (quantum) of acetylcholine. The quantal content refers to the number of vesicle-loads of acetylcholine released by the nerve terminal for each nerve impulse. Thus, the EPP amplitude is roughly equal to the mEPP amplitude multiplied by the quantal content.\n\nActive immunisation of experimental animals against the affinity-purified AChR, passive transfer with rat- or mouse-derived mono-clonal antibodies specific for the AChR, or passive transfer of purified MG immunoglobulins containing high levels of AChR antibodies have all been informative6–8. Both passive transfer and active immunisation animal models result in a reduced postsynaptic response to acetylcholine (the neurotransmitter) measured as a reduction in the amplitude of the EPP and mEPPs (Figure 1, normal on left and MG on right). As an animal becomes more severely affected, the EPP naturally becomes smaller and may not reach threshold for generation of the muscle action potential. A progressive failure of the action potential in a subset of myasthenic muscle fibres can be detected as a decrement in the compound muscle action potential (CMAP) amplitude during repetitive stimulation of the nerve5.\n\nBelow, we provide an update and brief summary of the current understanding of these synaptic diseases including the pathogenic effects of AChR antibodies upon the motor endplate, and some less well-known aspects that have recently been reviewed in detail3,9–11. This will be followed by recent approaches to begin to unravel the factors responsible for the failure of immune tolerance that leads to autoreactivity in MG. Finally, recent progress in our understanding of how MuSK autoantibodies cause NMJ failure will be discussed in detail.\n\n\nMechanisms of AChR antibodies\n\nAChR autoantibodies are mainly of the IgG1 and 3 subtypes, and so they are divalent and complement activating2. Binding of these antibodies to AChRs results in activation of the classical complement pathway with assembly of the membrane attack complex (MAC). Calcium influx through the MAC causes local damage to the membrane, with release of AChR-containing membrane debris into the synaptic cleft11. The damaged postsynaptic membrane shows a diminished response to acetylcholine, as measured electrophysiologically (Figure 1) by reduced amplitudes of EPPs and mEPPs. Importantly, and not widely appreciated, the complement damage also causes a loss of voltage-gated sodium channels, which are located in the secondary folds, raising the threshold that the EPP must reach to trigger the muscle action potential12. Bivalent AChR IgG can also cross-link adjacent AChRs, increasing the normally slow rate of internalisation and lysosomal degradation of the AChRs (normal half-life around 10 days in mice) and resulting in a loss of AChRs even in the absence of complement attack (Figure 1)13. Surprisingly, perhaps, most of the antibodies do not cause direct block of AChR function, although AChR block has been shown with a few individual patient sera14.\n\nThere are many questions concerning the variability of muscle weakness between patients, and even within a patient. Some factors that could, in theory, contribute to this variability are the rate of diffusion of AChR antibodies from the serum into the very small synaptic cleft of each NMJ, the high number of the AChRs within this space that have to be targeted before a deficit in transmission occurs, and synaptic compensatory mechanisms that can be demonstrated in animal models. Regarding the latter, an increase in muscle AChR synthesis was found in passive IgG transfer experiments, and, similarly, increased mRNA for AChR subunits in biopsies from MG patients2, and an increase in the quantal content of acetylcholine released from the nerve terminal during each nerve impulse15. These adaptive responses would each tend to protect neuromuscular transmission from the pathogenic effects. The level of expression of tissue complement regulators could also influence the extent of NMJ damage11. This is particularly important given that complement attack damages both the AChR-containing membrane (reducing sensitivity to acetylcholine) and the number of voltage-gated sodium channels (raising the threshold for the muscle action potential), as mentioned above12. It seems likely that each of these modulating factors might differ between individuals and between muscles within an individual, explaining to some extent the variation in weakness and fatigue that is characteristic of all forms of MG.\n\n\nRecent approaches to investigating the failure of tolerance to AChR in MG\n\nMost work in this area uses experimental models of MG, usually terminal experimental autoimmune MG (EAMG). This can be induced by active immunisation against purified AChR from electric organs of the marine ray, Torpedo, or electric eel, with adjuvants6,16. Torpedo AChR can be purified at high concentrations and in large amounts, making it highly suitable for EAMG induction. Unfortunately, only a proportion of the Torpedo AChR antibodies cross-react with mouse AChR to induce disease, and adjuvants are considered necessary to break tolerance. Thus, although EAMG results have helped to throw light on NMJ defects, the relevance of any immunological findings must be considered carefully. In a recent series of experiments, transgenic interleukin (IL)-17-null mice confirmed previous findings of the importance of T-helper cells that express the pro-inflammatory cytokine IL-1717. Since IL-17 is also expressed by other types of immune cells, the authors used adoptive transfer of CD4+ T cells from either wild-type or IL-17-null mice to repopulate IL-17-deficient mice before trying to induce EAMG. Host mice that were repopulated with wild-type CD4+ T cells developed antibodies against the injected Torpedo AChR and subsequently also developed autoantibodies against murine AChR. This was accompanied by myasthenic weakness. Host mice populated with IL-17-/- CD4+ T cells developed similar levels of anti-Torpedo AChR but little anti-murine AChR and were resistant to EAMG17. The authors could not detect any CD4+ T cells autoreactive for the murine AChR α-subunit. The findings therefore suggest that Th-17 cells do not play a role in the immune response to xenogeneic AChR but that they may facilitate the breaking of self-tolerance to the mouse (self) AChR. The cellular mechanisms involved remain to be defined.\n\nPatients with AChR MG fall into three main categories: early onset MG (predominantly women <50), late onset MG (more frequently men over 50), and MG associated with thymoma. Since these early and late onset groups differ in their human leukocyte antigen (HLA) associations, and in their thymic pathology, but not their IgG AChR antibody characteristics18, the distinctive clinical and aetiological characteristics suggest that the autoantibodies may arise via distinct pathogenic mechanisms operating within these different patient groupings19,20. The breaking of tolerance in early onset MG appears to involve the thymus, either primarily or secondarily, but human cellular studies have so far failed to identify the defects involved in antibody production.\n\nRecent genome-wide association studies (GWAS) are making it possible to begin to dissect genetic predisposing factors for specific patient groups in MG. A GWAS of 649 early onset AChR MG patients from Northern Europe confirmed associations of AChR MG with the HLA class 1 region (specifically HLA-B*08) and with the ‘Protein Tyrosine Phosphatase, Non-Receptor Type 22’ (PTPN22) gene21. The same study identified a novel association with the ‘TNFAIP3-interacting protein 1’ (TNIP1) gene. A more recent GWAS of 1032 white North American AChR patients revealed both similarities and differences between the early onset and late onset AChR MG patient groups22. Both groups were associated with the HLA class 2 locus (albeit with distinct haplotypes) and with the ‘cytotoxic T-lymphocyte–associated protein 4’ gene (CTLA4, a T cell membrane protein previously implicated in autoimmune diseases). The late onset MG group specifically showed a strong association with ‘tumour necrosis factor receptor 4 superfamily, member 11a, NF-κB activator’ (TNFRSF11A), which encodes a protein involved in interactions between dendritic cells and T cells22. These studies have begun to identify factors that might help to explain the early and late onset aetiologies. Additional, larger GWASs might allow dissection of distinct genes, alleles, and pathogenic mechanisms for different subsets of MG patients and could be particularly interesting with respect to the late onset MG patients who now represent a much higher proportion of the total23.\n\n\nMechanisms of MuSK antibodies\n\nAChR MG is an immune-mediated disease with most of the effects dependent on the particular characteristics of the IgG antibodies. By contrast, MuSK MG appears to be principally a ‘pharmacological’ disease, where antibodies act to interfere directly with physiological mechanisms.\n\n\nMuSK IgG4 blocks MuSK signalling\n\nAnimal experiments show that MuSK IgG can cause MG. Mice that received repeated daily injections of patient IgG showed impaired neuromuscular transmission, with reductions in endplate AChR and in EPP amplitudes24–30. Similar changes to endplates were reported in mice, rats, and rabbits that were actively immunised with MuSK29,31–36. Most of the MuSK in MG patient plasma is of the IgG4 subtype, with relatively low titres for IgG1-337,38. This is interesting because the IgG4 subclass lacks the complement-activating properties of IgG1 and is considered functionally monovalent39, eliminating the two main pathogenic mechanisms of AChR MG. When the IgG4 and IgG1-3 fractions of MuSK patient IgG were separately injected into mice, the IgG4 fraction caused MG27, while the IgG1-3 (but not with an equivalent amount of MuSK antibodies) did not. In the active immunisation model, complement-deficient mice that were immunised against MuSK developed MG that was even more severe than complement-sufficient strains35. Thus, endplate damage by MuSK antibody does not appear to rely upon the classical immunopathology nor, because of lack of cross-linking, antigenic modulation mechanisms that drive AChR MG pathology. Furthermore, in the active and passive mouse models of AChR and MuSK MG, postsynaptic AChRs and the mEPPs were reduced to a similar extent but in the MuSK MG models there was no adaptive increase in the number of quanta of acetylcholine released by the nerve terminal27–29,35,36. Perhaps failure of presynaptic compensation explains why MuSK MG mice were weaker and MuSK MG patients are often more severely affected compared to AChR MG patients. The proposed effect of MuSK autoantibodies upon the mechanisms of postsynaptic differentiation and synaptic function is illustrated in Figure 2.\n\n(A) Healthy MuSK-mediated postsynaptic differentiation pathway at the neuromuscular junction (NMJ). Neural agrin secreted by the motor nerve terminal binds to LRP4, low-density lipoprotein receptor-related protein 4 (LRP4), which causes the dimerisation of MuSK. MuSK dimerisation causes phosphorylation of MuSK and associated proteins of the MuSK pathway, including Dok7 and the acetylcholine receptor (AChR) β-subunit. Rapsyn is recruited to the phosphorylated AChRs, stabilising postsynaptic clusters of AChRs. (B) Impaired postsynaptic differentiation in animal models of MuSK myasthenia gravis. MuSK autoantibodies are mainly of the immunoglobulin (Ig)G4 subclass. They block the assembly of the agrin-LRP4-MuSK complex. Interruption of MuSK kinase signalling leads to slow disassembly of the postsynaptic AChR clusters. A resultant decline in miniature endplate potential (mEPP) and EPP amplitude (not shown) results in failure of the muscle action potential and fatiguing weakness. Co-existing IgG1-3 antibodies, although lower concentration, may contribute but their pathogenic roles are not yet well defined. The compensatory presynaptic upregulation of quantal release found in AChR MG does not occur in MuSK MG.\n\nMuSK is found in the postsynaptic membrane of the NMJ, together with AChR40. The protein tyrosine kinase function of MuSK is activated when agrin, a proteoglycan from the nerve terminal, binds to MuSK via the co-receptor ‘low-density lipoprotein receptor-related protein 4’ (LRP4)41–44. MG patient MuSK antibodies mainly bind the Ig-like regions in the MuSK ectodomain, thereby blocking assembly and activation of the agrin-LRP4-MuSK complex. This explains why agrin-induced AChR clustering in the C2C12 cell model was inhibited by incubation in MuSK MG sera and IgG preparations45–47. In mice injected with MuSK MG IgG, a reduction in postsynaptic tyrosine phosphorylation was associated with accelerated loss of AChRs from the postsynaptic AChR cluster30,48, culminating in failure of neuromuscular transmission28. Thus, a combination of cell culture and mouse studies suggests that MuSK autoantibodies, which are mainly of the IgG4 type, block the natural activation of MuSK, leading to progressive loss of AChRs from the motor endplate and synaptic failure.\n\nHowever, this may not be the whole story. Both the IgG4 and IgG1-3 fractions of MuSK MG plasma were able to inhibit agrin-induced AChR clustering when added to C2C12 muscle cell cultures. The intracellular protein Dok7 binds and stabilises the MuSK dimer, thereby enhancing MuSK’s tyrosine kinase activity49. In a modified C2C12 model, AChR clustering was artificially induced by overexpressing Dok7. Despite the absence of agrin from this experimental system, both the IgG4 and IgG1-3 fractions still caused dispersal of the AChR clusters, suggesting that both IgG4 and IgG1-3 may affect MuSK independent of the interaction with LRP445. Since IgG1-3 MuSK antibodies might also activate complement, it is too early to say that this IgG subclass plays no role. Conceivably, MuSK IgG1-3 antibodies might selectively affect certain muscle groups, for example those with especially high expression of MuSK50, or where tissue complement regulators are deficient.\n\nAt healthy NMJs, there is a balance between clustering and cluster dispersal mechanisms. During embryonic development, and subsequently in mature muscle, MuSK functions to aggregate AChRs under the incoming motor nerve but, at the same time, acetylcholine released from the motor nerve terminal and acting upon these AChRs tends to dismantle AChR clusters51,52. It is thought that calcium influx through the AChR channel may be amplified by subsynaptic IP3 receptors53, activating calcium-dependent proteases that then trigger the internalisation and degradation of AChRs, reducing AChR clusters. At healthy NMJs, synapse formation and synapse disassembly are balanced54,55. Impaired MuSK signalling in MuSK MG would disrupt this balance. This has clinical implications. Cholinesterase inhibitors, such as pyridostigmine, are a first-line treatment for MG. They prolong the activation of endplate AChRs and thereby restore the EPP amplitude. However, in MuSK MG patients, they are often not helpful or not tolerated56. In the mouse passive IgG transfer model of MuSK MG (where MuSK signalling is inhibited), pyridostigmine was found to exacerbate endplate AChR loss and NMJ failure57, probably by increasing and prolonging the dismantling action of acetylcholine on AChRs.\n\n\nWhittling down the ‘seronegative’ cases\n\nA substantial fraction of MG patients reveal no detectable AChR or MuSK antibodies using the standard clinical radio-immunoprecipitation assays. Sensitive cell-based assays (CBAs) have recently shown that many of these ‘seronegative’ patients do indeed possess autoantibodies. These CBAs use fluorescently conjugated anti-human IgG to probe for patient antibodies binding to closely packed synaptic membrane proteins expressed on transfected cells. The CBAs can detect antibodies that recognise AChRs only when closely packed together, mimicking the close AChR packing at the endplate58,59. Close AChR packing may allow these antibodies to form stable divalent binding interactions, which are not possible in solution owing to the low concentration of AChRs. The AChR antibodies detected by CBA were mainly of the complement-fixing IgG1 subtype, similar to other AChR MG antibodies, and were able to passively transfer electrophysiological evidence of MG to mice58,60.\n\nOther studies found that some double seronegative MG patients possessed LRP4 antibodies (mainly IgG1 and IgG2)61–65. Clearly antibodies to LRP4 could be pathogenic, and animals immunised against LRP4 demonstrate myasthenic weakness with impairment of neuromuscular transmission in mice66, but the frequency of LRP4 antibodies has been variable. Antibodies to the secreted protein agrin, which is responsible for activating the LRP4/MuSK pathway, have been detected in small numbers of MG patients. However, most of the cases reported so far (10/12) also had antibodies to MuSK, LRP4, and/or AChR, and only two patients had no other antibodies detected67,68. The clinical and pathogenic significance of both LRP4 and agrin autoantibodies requires further investigation.\n\n\nConclusions\n\nDifferent subsets of MG patients develop autoantibodies with distinct target specificities, isotypes, and pathogenic mechanisms. Different pathogenic mechanisms then converge to cause loss of postsynaptic AChRs and increasing failure of neuromuscular transmission. This raises the need to investigate the immunological abnormalities specific to each of these categories of MG (as well as any common factors or pathways that might offer parsimonious therapeutic targets). The relative rarity of MuSK MG patients may make GWAS difficult, but the intriguing variation in the number of patients affected at different latitudes in the northern hemisphere (A. Vincent, unpublished data) raises the possibility of environmental factors contributing to disease aetiology. Mice actively immunised with MuSK generated a response characterised by IgG1 (which has characteristics similar to human IgG4), IL-4, and IL-10, analogous to the MuSK immunology found in MuSK MG patients32,35,69, suggesting that there is something about the antigen itself that determines the immunological characteristics. Perhaps this mouse model will be useful for studying how and why IgG4 antibodies to MuSK arise.\n\nRecent studies in MuSK MG have also focused attention on the molecular defences of the target organ: the NMJ. Local complement regulator proteins help protect the motor endplate from MAC-mediated damage in AChR MG70,71. Agrin/MuSK signalling provides a more general adaptive/protective response whenever there is a challenge to the function of the NMJ72. Overexpression of MuSK or the intracellular MuSK-activator protein DOK7 protected muscles against NMJ impairment in transgenic mouse models of several neuromuscular diseases73,74. On the other hand, the NMJs of people carrying hypomorphic alleles for MuSK-pathway genes75 might be more susceptible to AChR autoantibodies. Similarly, any hyper-activation of the postsynaptic IP3R1 receptor/calpain/caspase/CDK5 pathway52–55 conceivably might exacerbate the loss of postsynaptic AChR in AChR MG. These synapse-regulatory pathways offer potential targets for therapeutic interventions to ameliorate motor endplate damage in MG.\n\nSome of the studies in animal models of MuSK MG reported changes in nerve terminal structure and/or presynaptic transmitter release24,33,35. The presynaptic changes appear less robust than the postsynaptic changes. Nevertheless, the adaptive increase in presynaptic acetylcholine release that regularly occurs in models of AChR MG and in AChR MG patients15 failed in models of MuSK MG. These findings suggest that MuSK signalling may help to mediate the presynaptic adaptive response. Ideally, some of the findings should be confirmed in patient muscle biopsies, particularly the most affected bulbar or facial muscles, but this remains a considerable challenge.\n\n\nAbbreviations\n\nAChR, acetylcholine receptor; CBAs, cell-based assays; CMAP, compound muscle action potential; CTLA4, cytotoxic T-lymphocyte–associated protein 4; EPP, endplate potential; GWAS, genome-wide association study; HLA, human leukocyte antigen; Ig, immunoglobulin; IL-17, interleukin-17; LRP4, low-density lipoprotein receptor-related protein 4; MAC, membrane attack complex; MuSK, muscle-specific kinase; MG, myasthenia gravis; NMJ, neuromuscular junction; PTPN22, Protein Tyrosine Phosphatase, Non-Receptor Type 22; TNIP1, TNFAIP3-interacting protein 1; TNFRSF11A, tumour necrosis factor receptor 4 superfamily, member 11a, NF-κB activator.",
"appendix": "Competing interests\n\n\n\nThe University of Oxford holds a patent for MuSK antibody detection, licensed in the USA to Athena Diagnostics, and Angela Vincent receives a proportion of royalties. William D. Phillips has no disclosures.\n\n\nGrant information\n\nWilliam D. Phillips was supported by grants from NHMRC (570930) and MDA (MDA4172) and University of Sydney grant (William D. Phillips). Work on myasthenic syndromes in Oxford is supported by the Watney Trust, Myaware, the NIHR Oxford Biomedical Research Centre, and the Muscular Dystrophy Campaign.\n\n\nReferences\n\nBerrih-Aknin S, Frenkian-Cuvelier M, Eymard B: Diagnostic and clinical classification of autoimmune myasthenia gravis. J Autoimmun. 2014; 48–49: 143–148. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nVincent A: Unravelling the pathogenesis of myasthenia gravis. Nat Rev Immunol. 2002; 2(10): 797–804. PubMed Abstract | Publisher Full Text\n\nVerschuuren JJ, Huijbers MG, Plomp JJ, et al.: Pathophysiology of myasthenia gravis with antibodies to the acetylcholine receptor, muscle-specific kinase and low-density lipoprotein receptor-related protein 4. Autoimmun Rev. 2013; 12(9): 918–923. PubMed Abstract | Publisher Full Text\n\nPhillips WD, Christadoss P, Losen M, et al.: Guidelines for pre-clinical animal and cellular models of MuSK-myasthenia gravis. Exp Neurol. 2015; 270: 29–40. PubMed Abstract | Publisher Full Text\n\nPlomp JJ, Morsch M, Phillips WD, et al.: Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models. Exp Neurol. 2015; 270: 41–54. PubMed Abstract | Publisher Full Text\n\nPatrick J, Lindstrom J: Autoimmune response to acetylcholine receptor. Science. 1973; 180(4088): 871–2. PubMed Abstract | Publisher Full Text\n\nToyka KV, Brachman DB, Pestronk A, et al.: Myasthenia gravis: passive transfer from man to mouse. Science. 1975; 190(4212): 397–399. PubMed Abstract | Publisher Full Text\n\nKusner LL, Losen M, Vincent A, et al.: Guidelines for pre-clinical assessment of the acetylcholine receptor--specific passive transfer myasthenia gravis model-Recommendations for methods and experimental designs. Exp Neurol. 2015; 270: 3–10. PubMed Abstract | Publisher Full Text\n\nGomez AM, Van Den Broeck J, Vrolix K, et al.: Antibody effector mechanisms in myasthenia gravis-pathogenesis at the neuromuscular junction. Autoimmunity. 2010; 43(5–6): 353–370. PubMed Abstract | Publisher Full Text\n\nBaggi F, Antozzi C, Toscani C, et al.: Acetylcholine receptor-induced experimental myasthenia gravis: what have we learned from animal models after three decades? Arch Immunol Ther Exp (Warsz). 2012; 60(1): 19–30. PubMed Abstract | Publisher Full Text\n\nTüzün E, Christadoss P: Complement associated pathogenic mechanisms in myasthenia gravis. Autoimmun Rev. 2013; 12(9): 904–911. PubMed Abstract | Publisher Full Text\n\nRuff RL, Lennon VA: How myasthenia gravis alters the safety factor for neuromuscular transmission. J Neuroimmunol. 2008; 201–202: 13–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoutrari H, Kokla A, Tzartos SJ: Passive transfer of experimental myasthenia gravis via antigenic modulation of acetylcholine receptor. Eur J Immunol. 1992; 22(9): 2449–2452. PubMed Abstract | Publisher Full Text\n\nWhiting PJ, Vincent A, Newsom-Davis J: Acetylcholine receptor antibody characteristics in myasthenia gravis. Fractionation of alpha-bungarotoxin binding site antibodies and their relationship to IgG subclass. J Neuroimmunol. 1983; 5(1): 1–9. PubMed Abstract | Publisher Full Text\n\nPlomp JJ, Van Kempen GT, DeBaets M, et al.: Acetylcholine release in myasthenia gravis: regulation at single end-plate level. Ann Neurol. 1995; 37(5): 627–636. PubMed Abstract | Publisher Full Text\n\nWu B, Goluszko E, Huda R, et al.: Experimental autoimmune myasthenia gravis in the mouse. In Curr Protoc Immunol. John Wiley & Sons, Inc. 2013; Chapter 15:unit 15.8. PubMed Abstract | Publisher Full Text\n\nSchaffert H, Pelz A, Saxena A, et al.: IL-17-producing CD4+ T cells contribute to the loss of B-cell tolerance in experimental autoimmune myasthenia gravis. Eur J Immunol. 2015; 45(5): 1339–1347. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nCompston DA, Vincent A, Newsom-Davis J, et al.: Clinical, pathological, HLA antigen and immunological evidence for disease heterogeneity in myasthenia gravis. Brain. 1980; 103(3): 579–601. PubMed Abstract | Publisher Full Text\n\nBerrih-Aknin S, Le Panse R: Myasthenia gravis: a comprehensive review of immune dysregulation and etiological mechanisms. J Autoimmun. 2014; 52: 90–100. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMarx A, Porubsky S, Belharazem D, et al.: Thymoma related myasthenia gravis in humans and potential animal models. Exp Neurol. 2015; 270: 55–65. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGregersen PG, Kosoy R, Annette T, et al.: Risk for myasthenia gravis maps to a 151Pro→Ala Change in TNIP1 and to Human Leukocyte Antigen-B*08. Ann Neurol. 2012; 72(6): 927–935. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRenton AE, Pliner HA, Provenzano C, et al.: A genome-wide association study of myasthenia gravis. JAMA Neurol. 2015; 72(4): 396–404. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nVincent A, Rothwell P: Myasthenia gravis. Autoimmunity. 2004; 37: 317–9. PubMed Abstract | Publisher Full Text\n\nCole RN, Reddel SW, Gervásio OL, et al.: Anti-MuSK patient antibodies disrupt the mouse neuromuscular junction. Ann Neurol. 2008; 63(6): 782–789. PubMed Abstract | Publisher Full Text\n\nCole RN, Ghazanfari N, Ngo ST, et al.: Patient autoantibodies deplete postsynaptic muscle-specific kinase leading to disassembly of the ACh receptor scaffold and myasthenia gravis in mice. J Physiol. 2010; 588(Pt 17): 3217–3229. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKawakami Y, Ito M, Hirayama M, et al.: Anti-MuSK autoantibodies block binding of collagen Q to MuSK. Neurology. 2011; 77(20): 1819–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlooster R, Plomp JJ, Huijbers MG, et al.: Muscle-specific kinase myasthenia gravis IgG4 autoantibodies cause severe neuromuscular junction dysfunction in mice. Brain. 2012; 135(Pt 4): 1081–1101. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMorsch M, Reddel SW, Ghazanfari N, et al.: Muscle specific kinase autoantibodies cause synaptic failure through progressive wastage of postsynaptic acetylcholine receptors. Exp Neurol. 2012; 237(2): 286–295. PubMed Abstract | Publisher Full Text\n\nViegas S, Jacobson L, Waters P, et al.: Passive and active immunization models of MuSK-Ab positive myasthenia: Electrophysiological evidence for pre and postsynaptic defects. Exp Neurol. 2012; 234(2): 506–512. PubMed Abstract | Publisher Full Text\n\nGhazanfari N, Morsch M, Reddel SW, et al.: Muscle-specific kinase (MuSK) autoantibodies suppress the MuSK pathway and ACh receptor retention at the mouse neuromuscular junction. J Physiol. 2014; 592(13): 2881–2897. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShigemoto K, Kubo S, Maruyama N, et al.: Induction of myasthenia by immunization against muscle-specific kinase. J Clin Invest. 2006; 116(4): 1016–1024. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJha S, Xu K, Maruta T, et al.: Myasthenia gravis induced in mice by immunization with the recombinant extracellular domain of rat muscle-specific kinase (MuSK). J Neuroimmunol. 2006; 175(1–2): 107–117. PubMed Abstract | Publisher Full Text\n\nPunga AR, Lin S, Oliveri F, et al.: Muscle-selective synaptic disassembly and reorganization in MuSK antibody positive MG mice. Exp Neurol. 2011; 230(2): 207–217. PubMed Abstract | Publisher Full Text\n\nRichman DP, Nishi K, Morell SW, et al.: Acute severe animal model of anti-muscle-specific kinase myasthenia: combined postsynaptic and presynaptic changes. Arch Neurol. 2012; 69(4): 453–460. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMori S, Kubo S, Akiyoshi T, et al.: Antibodies against muscle-specific kinase impair both presynaptic and postsynaptic functions in a murine model of myasthenia gravis. Am J Pathol. 2012; 180(2): 798–810. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPatel V, Oh O, Voit A, et al.: Altered active zones, vesicle pools, nerve terminal conductivity, and morphology during experimental MuSK myasthenia gravis. PLoS One. 2014; 9(12): e110571. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHoch W, McConville J, Helms S, et al.: Auto-antibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies. Nat Med. 2001; 7(3): 365–368. PubMed Abstract | Publisher Full Text\n\nMcConville J, Farrugia ME, Beeson D, et al.: Detection and characterization of MuSK antibodies in seronegative myasthenia gravis. Ann Neurol. 2004; 55(4): 580–584. PubMed Abstract | Publisher Full Text\n\nTao MH, Canfield SM, Morrison SL: The differential ability of human IgG1 and IgG4 to activate complement is determined by the COOH-terminal sequence of the CH2 domain. J Exp Med. 1991; 173(4): 1025–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nValenzuela DM, Stitt TN, DiStefano PS, et al.: Receptor tyrosine kinase specific for the skeletal muscle lineage: expression in embryonic muscle, at the neuromuscular junction, and after injury. Neuron. 1995; 15(3): 573–84. PubMed Abstract | Publisher Full Text\n\nDeChiara TM, Bowen DC, Valenzuela DM, et al.: The receptor tyrosine kinase MuSK is required for neuromuscular junction formation in vivo. Cell. 1996; 85(4): 501–512. PubMed Abstract | Publisher Full Text\n\nKim N, Stiegler AL, Cameron TO, et al.: Lrp4 is a receptor for Agrin and forms a complex with MuSK. Cell. 2008; 135(2): 334–342. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nZhang B, Luo S, Wang Q, et al.: LRP4 serves as a coreceptor of agrin. Neuron. 2008; 60(2): 285–297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhazanfari N, Fernandez KJ, Murata Y, et al.: Muscle specific kinase: organiser of synaptic membrane domains. Int J Biochem Cell Biol. 2011; 43(3): 295–298. PubMed Abstract | Publisher Full Text\n\nKoneczny I, Cossins J, Waters P, et al.: MuSK myasthenia gravis IgG4 disrupts the interaction of LRP4 with MuSK but both IgG4 and IgG1-3 can disperse preformed agrin-independent AChR clusters. PLoS One. 2013; 8(11): e80695. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuijbers MG, Zhang W, Klooster R, et al.: MuSK IgG4 autoantibodies cause myasthenia gravis by inhibiting binding between MuSK and Lrp4. Proc Natl Acad Sci U S A. 2013; 110(51): 20783–20788. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nOtsuka K, Ito M, Ohkawara B, et al.: Collagen Q and anti-MuSK autoantibody competitively suppress agrin/LRP4/MuSK signaling. Sci Rep. 2015; 5: 13928. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGhazanfari N, Linsao EL, Trajanovska S, et al.: Forced expression of muscle specific kinase slows postsynaptic acetylcholine receptor loss in a mouse model of MuSK myasthenia gravis. Physiol Rep. 2015; 3(12): pii: e12658. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBergamin E, Hallock PT, Burden SJ, et al.: The cytoplasmic adaptor protein Dok7 activates the receptor tyrosine kinase MuSK via dimerization. Mol Cell. 2010; 39(1): 100–109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPunga AR, Maj M, Lin S, et al.: MuSK levels differ between adult skeletal muscles and influence postsynaptic plasticity. Eur J Neurosci. 2011; 33(5): 890–8. PubMed Abstract | Publisher Full Text\n\nMisgeld T, Kummer TT, Lichtman JW, et al.: Agrin promotes synaptic differentiation by counteracting an inhibitory effect of neurotransmitter. Proc Natl Acad Sci U S A. 2005; 102(31): 11088–11093. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin W, Dominguez B, Yang J, et al.: Neurotransmitter acetylcholine negatively regulates neuromuscular synapse formation by a Cdk5-dependent mechanism. Neuron. 2005; 46(4): 569–579. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nZhu H, Bhattacharyya BJ, Lin H, et al.: Skeletal muscle IP3R1 receptors amplify physiological and pathological synaptic calcium signals. J Neurosci. 2011; 31(43): 15269–15283. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen F, Qian L, Yang ZH, et al.: Rapsyn interaction with calpain stabilizes AChR clusters at the neuromuscular junction. Neuron. 2007; 55(2): 247–260. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWang JY, Chen F, Fu XQ, et al.: Caspase-3 cleavage of dishevelled induces elimination of postsynaptic structures. Dev Cell. 2014; 28(6): 670–84. PubMed Abstract | Publisher Full Text\n\nEvoli A, Padua L: Diagnosis and therapy of myasthenia gravis with antibodies to muscle-specific kinase. Autoimmun Rev. 2013; 12(9): 931–5. PubMed Abstract | Publisher Full Text\n\nMorsch M, Reddel SW, Ghazanfari N, et al.: Pyridostigmine but not 3,4-diaminopyridine exacerbates ACh receptor loss and myasthenia induced in mice by muscle-specific kinase autoantibody. J Physiol (Lond). 2013; 591(10): 2747–2762. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeite M, Jacob S, Viegas S, et al.: IgG1 antibodies to acetylcholine receptors in 'seronegative' myasthenia gravis. Brain. 2008; 131(Pt 7): 1940–52. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nDevic P, Petiot P, Simonet T, et al.: Antibodies to clustered acetylcholine receptor: expanding the phenotype. Eur J Neurol. 2014; 21(1): 130–4. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nJacob S, Viegas S, Leite MI, et al.: Presence and pathogenic relevance of antibodies to clustered acetylcholine receptor in ocular and generalized myasthenia gravis. Arch Neurol. 2012; 69(8) 994–1001. PubMed Abstract | Publisher Full Text\n\nHiguchi O, Hamuro J, Motomura M, et al.: Autoantibodies to low-density lipoprotein receptor-related protein 4 in myasthenia gravis. Ann Neurol. 2011; 69(2): 418–422. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPevzner A, Schoser B, Peters K, et al.: Anti-LRP4 autoantibodies in AChR- and MuSK-antibody-negative myasthenia gravis. J Neurol. 2012; 259(3): 427–435. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nZhang B, Tzartos JS, Belimezi M, et al.: Autoantibodies to lipoprotein-related protein 4 in patients with double-seronegative myasthenia gravis. Arch Neurol. 2012; 69(4): 445–51. PubMed Abstract | Publisher Full Text\n\nCossins J, Belaya K, Zoltowska K, et al.: The search for new antigenic targets in myasthenia gravis. Ann N Y Acad Sci. 2012; 1275: 123–8. PubMed Abstract | Publisher Full Text\n\nZisimopoulou P, Evangelakou P, Tzartos J, et al.: A comprehensive analysis of the epidemiology and clinical characteristics of anti-LRP4 in myasthenia gravis. J Autoimmun. 2014; 52: 139–45. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nShen C, Lu Y, Zhang B, et al.: Antibodies against low-density lipoprotein receptor-related protein 4 induce myasthenia gravis. J Clin Invest. 2013; 123(12): 5190–202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGasperi C, Melms A, Schoser B, et al.: Anti-agrin autoantibodies in myasthenia gravis. Neurology. 2014; 82(22): 1976–83. PubMed Abstract | Publisher Full Text\n\nZhang B, Shen C, Bealmear B, et al.: Autoantibodies to agrin in myasthenia gravis patients. PLoS One. 2014; 9(3): e91816. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUlusoy C, Kim E, Tüzün E, et al.: Preferential production of IgG1, IL-4 and IL-10 in MuSK-immunized mice. Clin Immunol. 2014; 151(2): 155–163. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMorgan BP, Chamberlain-Banoub J, Neal JW, et al.: The membrane attack pathway of complement drives pathology in passively induced experimental autoimmune myasthenia gravis in mice. Clin Exp Immunol. 2006; 146(2): 294–302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKusner LL, Halperin JA, Kaminski HJ: Cell surface complement regulators moderate experimental myasthenia gravis pathology. Muscle Nerve. 2013; 47(1): 33–40. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSamuel MA, Valdez G, Tapia JC, et al.: Agrin and synaptic laminin are required to maintain adult neuromuscular junctions. PLos One. 2012; 7(10): e46663. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPérez-García MJ, Burden SJ: Increasing MuSK activity delays denervation and improves motor function in ALS mice. Cell rep. 2012; 2(3): 497–502. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nArimura S, Okada T, Tezuka T, et al.: Neuromuscular disease. DOK7 gene therapy benefits mouse models of diseases characterized by defects in the neuromuscular junction. Science. 2014; 345(6203): 1505–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nEngel AG, Shen XM, Selcen D, et al.: What have we learned from the congenital myasthenic syndromes. J Mol Neurosci. 2010; 40(1–2): 143–153. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "14609",
"date": "27 Jun 2016",
"name": "Lin Mei",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14610",
"date": "27 Jun 2016",
"name": "Marc De Baets",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1513
|
https://f1000research.com/articles/5-762/v1
|
27 Apr 16
|
{
"type": "Research Article",
"title": "Combined analysis of trabectome and phaco-trabectome outcomes by glaucoma severity",
"authors": [
"Yalong Dang",
"Pritha Roy",
"Igor I. Bussel",
"Ralitsa T. Loewen",
"Hardik Parikh",
"Nils A. Loewen",
"Yalong Dang",
"Pritha Roy",
"Igor I. Bussel",
"Ralitsa T. Loewen",
"Hardik Parikh"
],
"abstract": "Prior glaucoma severity staging systems were mostly concerned with visual field function and retinal nerve fiber layer, but did not include intraocular pressure or medications to capture resistance to treatment. We recently introduced a simple index that combines pressure, medications, and visual field damage and applied it to stratify outcomes of trabectome surgery. This microincisional glaucoma surgery removes the primary resistance to outflow in glaucoma, the trabecular meshwork, and has been mostly used in mild glaucoma. Traditional glaucoma surgeries have a relatively high complication rate and have been reserved for more advanced disease stages. In the analysis presented here we include our data of trabectome combined with cataract surgery. This is a common practice pattern as both occur in the same age group with increasing frequency. For patients in higher glaucoma index (GI) groups, the intraocular pressure (IOP) reduction was 2.34+/-0.19 mmHg more than those in a GI group one level lower while holding everything else constant. Those who had undergone trabectome combined with phacoemulsification had an IOP reduction that was 1.29+/-0.39 mmHg less compared to those with trabectome alone. No statistically significant difference was found between genders and age groups while holding everything else constant. Hispanics had a 3.81+/-1.08 mmHg greater IOP reduction. Pseudoexfoliation and steroid glaucoma patients had an IOP reduction that was greater by 2.91+/-0.56 and 3.86+/-0.81 mmHg, respectively, than those with primary open angle glaucoma. These results suggest a role for trabectome-mediated ab interno trabeculectomy beyond mild forms of glaucoma. Additionally, the multifactorial glaucoma index demonstrates a role in staging patients when comparing glaucoma surgical modalities.",
"keywords": [
"glaucoma",
"outflow",
"surgery",
"trabectome",
"ab interno trabeculectomy",
"disease index"
],
"content": "Introduction\n\nDue to an increasing human lifespan, chronic diseases that manifest later in life, such as glaucoma and cataracts, have an increasing incidence and often occur in the same individuals1. In addition, medications used to treat glaucoma2 or interventions to reduce intraocular pressure (IOP)3 can cause cataracts or accelerate their occurrence. Traditional glaucoma surgery consists of trabeculectomy or tube shunt implantation, both of which have a relatively high frequency of serious complications. Ab interno trabeculectomy with the trabectome, a plasma surgical modality that ionizes and aspirates the trabecular meshwork with minimal energy transfer to surrounding tissues, was first introduced in 20044 and is considered a more mature microincisional glaucoma surgery. Trabectome surgery, similar to other surgeries in this family, has a favorable safety profile5 but is often only performed in ocular hypertension or mild glaucoma stages6.\n\nThe primary outflow resistance in glaucoma is the trabecular meshwork7. However, more recent insight has also demonstrated a significant contribution to elevated pressure in glaucoma by an outflow resistance that is downstream of the trabecular meshwork8–10. According to the Goldmann equation, the limiting factor in pressure reduction from ab interno trabeculectomy with the trabectome is this residual resistance and the pressure of the episcleral veins11, and this may vary depending on glaucoma type12 and severity. Glaucoma severity can be described by visual field13, optic nerve damage, and also by the number of medications needed to achieve a target IOP14.\n\nWe recently examined the amount of IOP reduction that is due to phacoemulsification at the time of trabectome surgery and found this to be relatively clinically insignificant9. We also noted that trabectome surgery performed after failed trabeculectomy caused patients with more advanced visual field damage to have on average a greater pressure reduction despite similar medications10. In the current study we have consequently combined data of trabectome and phaco-trabectome surgery patients and stratified them by a glaucoma severity index. By combining both trabectome and phaco-trabectome surgery data, a more complete picture can be obtained to guide surgeons on whether ab interno trabeculectomy may be an appropriate primary intervention.\n\n\nMethods\n\nThis retrospective analysis was approved (PRO14100026) by the Institutional Review Board of the University of Pittsburgh in accordance with the Declaration of Helsinki and the Health Insurance Portability and Accountability Act. Because of the retrospective nature, no consent was required. Glaucoma patients who received trabectome with or without phacoemulsification were enrolled, except in the following circumstances: history of glaucoma surgery, any subsequent cataract or glaucoma surgery in the follow-up period, and short term followup (less than 12 months). Patients were divided into four groups (from mild to severe) according to a glaucoma index (GI), an indicator of glaucoma severity based on visual field, numbers of glaucoma medication, and preoperative IOP14. GI group 1 = mild, GI group 2 = moderate, GI group 3 = advanced, and GI4 = severe were defined based on glaucoma index scores of “≤4”, “4<GI≤8”, “8<GI≤16”, and “>16,” respectively. The main outcome measure was the reduction of IOP. Secondary outcome measures included reduction of medication and a Kaplan-Meier survival analysis. Baseline characteristics were analyzed by the Kruskal-Wallis and chisquare tests for continuous and categorical variables between GI groups, respectively. Univariate linear regression was performed first and those demographics found to be statistically significant were included into the multivariate regression analysis. Kaplan-Meier was used for survival-curve analyses. Surgical success was defined as IOP≤21 mmHg or at least 20% IOP reduction from baseline in any two consecutive visits after three months and no secondary glaucoma surgery. Log-rank test was used to compare survival distributions of GI groups.\n\n\nResults\n\nA total of 1340 cases of glaucoma patients were enrolled in the study and most of them were primary open angle glaucoma (POAG). The distribution across glaucoma severity groups was relatively even in number and average ages (Table 1). There was a slight preponderance of female patients in the mild and moderate groups. The ethnicity of most patients was Caucasian followed by Asian. POAG and pseudoexfoliation glaucoma were the most common diagnoses. The cup disc ratio increased by glaucoma index group and more patients were phakic than pseudophakic. More patients in the higher GI groups had a trabectome surgery that was combined with cataract surgery. Patients with a higher GI group had a more profound IOP reduction (Figure 1). At one year, the mean IOP reduction was 3.57±5.01, 5.34±5.40, 7.75±7.40, 12.09±8.08 mmHg for GI group 1 to 4, respectively. This pressure decrease occurred already on day 1 and remained relatively stable (Figure 2). Similarly, patients with more severe glaucoma experienced a larger reduction in medications which were tapered more gradually (Figure 3). When we stratified the overall IOP reduction by glaucoma severity, patients with worse glaucoma had the largest decrease.\n\nGlaucoma index (GI); GI1 Mild: GI≤4; GI2 Moderate: 4<GI≤8; GI3 Advanced: 8<GI≤16; GI4 Severe: GI>16.\n\nMore severe glaucoma was associated with a larger pressure reduction.\n\nPatients with a higher group had the largest decrease (average and standard deviation).\n\nPatients in the severe and advanced groups had the largest medication reduction (average and standard deviation).\n\nIn the univariate regression analysis, age was slightly negatively correlated with the amount of IOP reduction (Table 2) but this was not noted in the multivariate regression (Table 3) while male gender had a positive correlation in the univariate but not anymore in the multivariate regression. For patients in the higher GI group, the IOP reduction was 2.34±0.19 mmHg more than those in one level lower GI group while holding everything else constant. Hispanics experienced a pressure drop larger by 3.81±1.08 mmHg than other ethnicities as did patients with a diagnosis of pseudoexfoliation and steroid induced glaucoma (Table 3). IOP reduction was 2.91±0.56 and 3.86±0.81 mmHg more than in POAG patients. Interestingly, cataract surgery was associated with a slightly worse IOP reduction by 1.29+/-0.39 mmHg (Table 3).\n\nSurvival rate at 12 months was 93%, 84%, 82% and 74% for GI group 1 to 4 (Figure 4). Log-rank test indicated statistically significant differences between the GI groups and patients in the lower GI groups had a higher survival rate than those in higher GI groups.\n\nThe highest GI group with more advanced glaucoma (GI4, red) had the worst survival.\n\n\nDiscussion\n\nThe results of the current study are confirmatory of our prior study where we examined the impact of a glaucoma severity index on the results of trabectome surgery when done as a standalone procedure14. The larger number of patients involved here allowed discovery of additional factors. We included here phaco-trabectome patients, who have a different, mixed indication that often includes visually significant cataract as the primary motivator while presenting with a relatively stable glaucoma. We did so after demonstrating by a rigorous statistical matching method, coarsened exact matching15, that phacoemulsification does not contribute significantly to IOP reduction when done at the same time9 or in a surgery prior to trabectome surgery16.\n\nThe results of this study match established risk factors and findings from other studies. Steroid glaucoma and pseudoexfoliation often produce very high IOPs and the primary pathology is located in the trabecular meshwork. As a result, ablating the meshwork reduces intraocular pressure very effectively17.\n\nIn that study we found that a larger pressure reduction is achieved in more severe glaucoma consisting of a more advanced visual field damage, a higher pre-intervention IOP and more medications. We had previously found that cup disc ratio, Hispanic ethnicity and diagnosis of steroid-induced glaucoma are related to a larger IOP reduction14. In addition to steroid induced glaucoma, pseudoexfoliation glaucoma confers a larger IOP reduction in the present study. Although phacoemulsification was negatively correlated with IOP reduction in this analysis, something that has been described for combined traditional trabeculectomy18,19, it is important to recall that this study was not designed to formally compare outcomes of combined versus trabectome-alone as we have done before9. The number of patients analyzed here is significantly higher than in the two separate studies but we note similar results in the regression analysis allowing to discover additional factors.\n\nThe Goldmann equation describes that the limiting factor to IOP reduction after removal of the trabecular meshwork, the substrate of the main outflow resistance, is the episcleral venous pressure and uveoscleral outflow11. The data presented here indicate that that this is mostly true also for more advanced glaucoma and consistent with two prior studies that demonstrated that significant conventional outflow can be recovered even after failed tube shunts20 and after failed trabeculectomy10. Small differences of the eventually achieved pressures could be explained by an episcleral venous pressures that is higher in glaucoma12. Overall, the data presented here suggest that ab interno trabeculectomy might be an appropriate surgery to attempt to control more than mild glaucoma.\n\n\nData availability\n\nThe raw datasets could not be made available because the data could not be sufficiently anonymised to protect patient confidentiality. No individuals other than the investigators or research staff involved in the conduct of this research study and authorized representatives of the University Research Conduct and Compliance Office (RCCO) are permitted access to research data or documents (including medical record information) associated with the conduct of this research study. Institutional IRB rules are available on the following University of Pittsburgh OSIRIS website: http://www.osiris.pitt.edu/osiris. The approval permit number for this study is PRO14100026.",
"appendix": "Author contributions\n\n\n\nYD, PR, RTL and NAL acquired and analyzed the data, all authors participated in writing and reviewing the manuscript, NAL provided funding.\n\n\nCompeting interests\n\n\n\nNAL has received honoraria for trabectome wet labs and lectures from Neomedix Corp.\n\n\nGrant information\n\nThis study was funded by the Eye and Ear Foundation of Pittsburgh.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nGuedes G, Tsai JC, Loewen NA: Glaucoma and aging. Curr Aging Sci. 2011; 4(2): 110–7. PubMed Abstract | Publisher Full Text\n\nChandrasekaran S, Cumming RG, Rochtchina E, et al.: Associations between elevated intraocular pressure and glaucoma, use of glaucoma medications, and 5-year incident cataract: the Blue Mountains Eye Study. Ophthalmology. 2006; 113(3): 417–24. PubMed Abstract | Publisher Full Text\n\nPatel HY, Danesh-Meyer HV: Incidence and management of cataract after glaucoma surgery. Curr Opin Ophthalmol. 2013; 24(1): 15–20. PubMed Abstract | Publisher Full Text\n\nMinckler DS, Baerveldt G, Alfaro MR, et al.: Clinical results with the Trabectome for treatment of open-angle glaucoma. Ophthalmology. 2005; 112(6): 962–7. PubMed Abstract | Publisher Full Text\n\nKaplowitz K, Bussel II, Honkanen R, et al.: Review and meta-analysis of ab-interno trabeculectomy outcomes. Br J Ophthalmol. 2016. pii: bjophthalmol-2015-307131. PubMed Abstract | Publisher Full Text\n\nKaplowitz K, Schuman JS, Loewen NA: Techniques and outcomes of minimally invasive trabecular ablation and bypass surgery. Br J Ophthalmol. 2014; 98(5): 579–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEthier CR, Kamm RD, Palaszewski BA, et al.: Calculations of flow resistance in the juxtacanalicular meshwork. Invest Ophthalmol Vis Sci. 1986; 27(12): 1741–50. PubMed Abstract\n\nSchuman JS, Chang W, Wang N, et al.: Excimer laser effects on outflow facility and outflow pathway morphology. Invest Ophthalmol Vis Sci. 1999; 40(8): 1676–80. PubMed Abstract\n\nParikh HA, Bussel II, Schuman JS, et al.: Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One. 2016; 11(2): e0149384. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBussel II, Kaplowitz K, Schuman JS, et al.: Outcomes of ab interno trabeculectomy with the trabectome after failed trabeculectomy. Br J Ophthalmol. BMJ Publishing Group Ltd. 2015; 99(2): 258–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrubaker RF: Goldmann’s equation and clinical measures of aqueous dynamics. Exp Eye Res. 2004; 78(3): 633–7. PubMed Abstract | Publisher Full Text\n\nSelbach JM, Posielek K, Steuhl KP, et al.: Episcleral venous pressure in untreated primary open-angle and normal-tension glaucoma. Ophthalmologica. 2005; 219(6): 357–61. PubMed Abstract | Publisher Full Text\n\nNg M, Sample PA, Pascual JP, et al.: Comparison of visual field severity classification systems for glaucoma. J Glaucoma. 2012; 21(8): 551–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoewen RT, Roy P, Parikh HA, et al.: Impact of a Glaucoma Severity Index on Results of Trabectome Surgery: Larger Pressure Reduction in More Severe Glaucoma. PLoS One. 2016; 11(3): e0151926. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIacus SM, King G, Porro G: Causal Inference without Balance Checking: Coarsened Exact Matching. Polit Anal. 2012; 20(1): 1–24. Publisher Full Text\n\nBussel I, Neiweem A, Schuman J, et al.: Glaucoma Surgery Calculator: Limited Additive IOP Effect of Phacoemulsification on Ab Interno Trabeculectomy. F1000Res. 2016; 5. [cited 2016 Mar 20]. Publisher Full Text\n\nWidder RA, Dinslage S, Rosentreter A, et al.: A new surgical triple procedure in pseudoexfoliation glaucoma using cataract surgery, Trabectome, and trabecular aspiration. Graefes Arch Clin Exp Ophthalmol. 2014; 252(12): 1971–5. PubMed Abstract | Publisher Full Text\n\nNaveh N, Kottass R, Glovinsky J, et al.: The long-term effect on intraocular pressure of a procedure combining trabeculectomy and cataract surgery, as compared with trabeculectomy alone. Ophthalmic Surg. 1990; 21(5): 339–45. PubMed Abstract\n\nKosmin AS, Wishart PK, Ridges PJ: Long-term intraocular pressure control after cataract extraction with trabeculectomy: phacoemulsification versus extracapsular technique. J Cataract Refract Surg. Elsevier; 1998; 24(2): 249–55. PubMed Abstract | Publisher Full Text\n\nMosaed S, Chak G, Haider A, et al.: Results of Trabectome Surgery Following Failed Glaucoma Tube Shunt Implantation: Cohort Study. Medicine (Baltimore). 2015; 94(30): e1045. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "13590",
"date": "10 May 2016",
"name": "Kristy G. Mascarenhas",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis was an interesting analysis of the use of Trabectome surgery in patients with more severe glaucomas. Comments and questions are as follows, most of which would clarify the methods and data analysis for the reader:The abstract and introduction could be clearer about the purpose of the study. You have to wait a long time to find out that this was a retrospective chart review of patients who had received Trabectome and phaco-Trabectome surgeries. Were the authors looking for statistically significant differences in IOP reduction between the glaucoma severity groups, or equivalent effect in mild vs. severe glaucoma? If so, was a power calculation done? The abstract makes it sound as though only phaco-Trabectome cases were included, but the introduction states that both Trabectome and combined surgeries were included. Please clarify the abstract? The methods section briefly introduces the glaucoma index system that was used, but especially for someone who has not read the previous article evaluating it, a bit more detail about the point system used would be helpful. One of the exclusion criteria was prior glaucoma surgery. Is data available about the number of patients who had previously had SLT/ALT? Table 1 shows the number of Trabectomes vs. phaco/Trabectomes. Did all phakic patients have cataract surgery, or did any receive Trabectome alone? Over what time period did the reviewed patients receive Trabectome? Relatively recently, or closer to the beginning of Trabectome use? How many different surgeons' patients were reviewed? (Is it possible that surgical techniques evolved over time, or that different surgeons' techniques varied significantly?) A sentence like \"We reviewed the charts of XXX patients from XXX surgeons who underwent CE/Trabectome or Trabectome alone between 2012-2016\" would be very helpful. The authors state that patients in the severe GI group and patients with more severe glaucomas like pseudoexfoliation and steroid-response had a greater IOP reduction than other groups. Since IOP is one component of the glaucoma index and Trabectome does lower IOP more significantly with higher pre-op IOP, is it possible to say that Trabectome had a greater effect in these groups independently of pre-op IOP? (To confirm, was this analysis done?) Can the authors include a short discussion of limitations of the current study and possibly a bit of speculation on the smaller IOP reduction with phaco/Trabectome than with Trabectome alone, as well as on why the more severe GI groups saw a shorter-lived IOP reduction?",
"responses": [
{
"c_id": "2018",
"date": "27 Jun 2016",
"name": "Nils Loewen",
"role": "Author Response",
"response": "This was an interesting analysis of the use of Trabectome surgery in patients with more severe glaucomas. Comments and questions are as follows, most of which would clarify the methods and data analysis for the reader: The abstract and introduction could be clearer about the purpose of the study. You have to wait a long time to find out that this was a retrospective chart review of patients who had received Trabectome and phaco-Trabectome surgeries. Were the authors looking for statistically significant differences in IOP reduction between the glaucoma severity groups, or equivalent effect in mild vs. severe glaucoma? If so, was a power calculation done? Authors: Thank you for the opportunity to revise our abstract and introduction. We have added to the abstract: “In the analysis presented here, we combined data of trabectome alone and trabectome with same session cataract surgery to increase testing power and chances of effect discovery.” We have also completely rewritten the introduction to explain the purpose of combining trabectome and phaco-trabectome data better and the idea behind the glaucoma index. Briefly, we recently created a simple glaucoma index that combines IOP, the number of medications and visual field status to gauge relative clinical glaucoma severity but also describes resistance to treatment. This may control for medication use when trying to gauge IOP results and allow to focus on glaucoma severity related outcomes. We combined trabectome and phaco-trabectome data sets because phacoemulsification is not a significant contributor to IOP reduction in studies with advanced matching strategies. A power calculation was done and our regression analysis has a power of 0.95. We would require 82 patients to detect a GI IOP difference for a medium sized effect of 0.3 and a power of 80%. As our statistician advises, a power analysis is only needed when no statistical significance is found and that is why we had not reported it in our original submission. If a statistically significant difference is found as here, then that suggests that the statistical test is powerful enough. The abstract makes it sound as though only phaco-Trabectome cases were included, but the introduction states that both Trabectome and combined surgeries were included. Please clarify the abstract? Authors: Thank you for noticing this. Both were included and we have added the following to the abstract: “In the analysis presented here, we combined data of trabectome alone and trabectome with same session cataract surgery to increase testing power and chances of effect discovery.” The methods section briefly introduces the glaucoma index system that was used, but especially for someone who has not read the previous article evaluating it, a bit more detail about the point system used would be helpful. Authors: We have updated our methods section to explain the individual factors in more detail: “Patients were divided into four groups (from mild to severe) according to a glaucoma index (GI), an indicator of glaucoma severity based on visual field, the number of glaucoma medications, and preoperative IOP1. Baseline IOP was divided into 4 groups, <20 mmHg, 20-29 mmHg, 30-39 mmHg, and above 40 and assigned with 1 to 4 points. Glaucoma medications (meds) were divided into 4 groups: ≤1, 2, 3, or ≥4, and assigned values of 1 to 4. Visual field (VF) was separated into 4 groups with points from 1 to 4: mild, moderate, advanced and end-stage. GI was then defined as VF*meds*IOP and separated into GI group 1 = mild, GI group 2 = moderate, GI group 3 = advanced, and GI4 = severe defined based on glaucoma index scores of “≤4”, “4<GI≤8”, “8<GI≤16”, and “>16,” respectively.” One of the exclusion criteria was prior glaucoma surgery. Is data available about the number of patients who had previously had SLT/ALT? Authors: Laser trabeculoplasty is the first line of treatment in our practice pattern, and this was not an exclusion criterion. Nearly all patients will have had this. Trabeculoplasty was not included as a factor in the analysis to reduce complexity. We have added to Methods: “A history of laser trabeculoplasty did not lead to exclusion.” Table 1 shows the number of Trabectomes vs. phaco/Trabectomes. Did all phakic patients have cataract surgery, or did any receive Trabectome alone? Authors: We have analyzed the effect of trabectome in phakic patients before in two different studies2,3 and found that even in narrow angles there does not seem to be a significant difference. Consequently, we included here all patients, including trabectome patients who remained phakic, to focus on the glaucoma index without further break down into subgroups. It is important to recall that differences in phaco versus no phaco are merely based on averages. Because of a mixed indication for these patients (many cataract surgery patients do not need a pressure reduction and merely want to reduce medications) and because no formal matching was done, one cannot conclude from this study that cataract surgery is causatively linked to a diminished IOP reduction when compared to trabectome-only patients. On the contrary, an IOP reduction in many of our phaco-trabectome patients is simply not necessary to the same degree. Over what time period did the reviewed patients receive Trabectome? Relatively recently, or closer to the beginning of Trabectome use? How many different surgeons' patients were reviewed? (Is it possible that surgical techniques evolved over time, or that different surgeons' techniques varied significantly?) A sentence like \"We reviewed the charts of XXX patients from XXX surgeons who underwent CE/Trabectome or Trabectome alone between 2012-2016\" would be very helpful. Authors: The study period included data from 2011 up until the end of 2015. We did not analyze outcomes by individual surgeons. We concede that a majority of surgeries was performed by the senior author of this study and after 2012. Unfortunately, is not easily possible to provide the exact breakdown due to how data is extracted from the EMR data cloud. Manual curation of thousands of patient operative reports would be necessary. the authors state that patients in the severe GI group and patients with more severe glaucomas like pseudoexfoliation and steroid-response had a greater IOP reduction than other groups. Since IOP is one component of the glaucoma index and Trabectome does lower IOP more significantly with higher pre-op IOP, is it possible to say that Trabectome had a greater effect in these groups independently of pre-op IOP? (To confirm, was this analysis done?) Authors: Thank you very much for this inquiry. Without conducting a prospective, randomized controlled study or applying an advanced matching strategy, it is not possible to detect a strong, likely causative, correlation between glaucoma type and the amount of resulting IOP drop. The multivariate regression analysis we have done here does show that pseudoexfoliation and steroid glaucoma are significantly associated with a larger IOP reduction. This would be expected since both are obvious trabecular meshwork diseases. It is not clear what the substrate is that prevents IOP from reaching the theoretical limit of episcleral venous pressure after trabecular ablation or bypass. The glaucoma index already captures resistance to medical treatment and these eyes require an IOP that is considerably lower than what can be achieved by trabecular ablation. This may reflect a tipping point.4 Can the authors include a short discussion of limitations of the current study and possibly a bit of speculation on the smaller IOP reduction with phaco/Trabectome than with Trabectome alone, as well as on why the more severe GI groups saw a shorter-lived IOP reduction? Authors: Thank you. We would like to add to the discussion the following: “This study has several limitations: to maximize patient number, we included all trabectome surgeries regardless of lens status or lens-cosurgery because our prior studies indicated that neither had a clinically relevant impact.2,3,5 This study was limited to only one year when follow-up and data integrity was most amenable to automated retrieval and analysis. As a retrospective analysis, this study cannot discover causality and is only able to advise that patients with more severe glaucoma had a similar postoperative IOP and a comparable reduction in medications as those with mild glaucoma.” We have also added a better explanation why phaco-trabectome patients have a lesser IOP reduction: “We included here phaco-trabectome patients, who have a different, mixed indication that often includes a visually significant cataract as the primary motivator while presenting with a relatively stable glaucoma without the need for pressure reduction but a motivation to reduce medications. We did so after demonstrating by a rigorous statistical matching method, coarsened exact matching6, that phacoemulsification does not contribute significantly to IOP reduction when done at the same time 7 or in a surgery prior to trabectome surgery 5. As in our prior studies 5, phaco-trabectome patients had a lower preoperative IOP. This study focused on stratification of outcomes by glaucoma index and did not apply advanced matching strategies that we applied elsewhere 5,7,8 to compare two groups. Hence, one cannot conclude from this study that cataract surgery is causatively linked to a diminished IOP reduction when compared to trabectome-only patients. On the contrary, an IOP reduction in many of our phaco-trabectome patients is simply not necessary to the same degree.” References used in this Reply to Reviewers Loewen, R. T. et al. Impact of a Glaucoma Severity Index on Results of Trabectome Surgery: Larger Pressure Reduction in More Severe Glaucoma. PLoS One 11, e0151926 (2016). Bussel, I. I., Kaplowitz, K., Schuman, J. S., Loewen, N. A. & Trabectome Study Group. Outcomes of ab interno trabeculectomy with the trabectome by degree of angle opening. Br. J. Ophthalmol. 99, 914–919 (2015). Parikh, H. A., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One 11, e0149384 (2016). Wollstein, G. et al. Retinal nerve fibre layer and visual function loss in glaucoma: the tipping point. Br. J. Ophthalmol. 96, 47–52 (2012). Neiweem, A. E., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Glaucoma Surgery Calculator: Limited Additive Effect of Phacoemulsification on Intraocular Pressure in Ab Interno Trabeculectomy. PLoS One 11, e0153585 (2016). Iacus, S. M., King, G. & Porro, G. Causal Inference without Balance Checking: Coarsened Exact Matching. Polit. Anal. 20, 1–24 (2012). Parikh, H. A., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One 11, e0149384 (2016). Akil, H. et al. Clinical Results of Ab Interno Trabeculotomy Using the Trabectome in Patients with Pigmentary Glaucoma compared to Primary Open Angle Glaucoma. Clin. Experiment. Ophthalmol. (2016). doi:10.1111/ceo.12737 Gedde, S. J. et al. in Am J Ophthalmol 153, 804–814.e1 (2012 Elsevier Inc, 2012). Gedde, S. J. et al. Treatment Outcomes in the Tube Versus Trabeculectomy (TVT) Study After Five Years of Follow-up. Am. J. Ophthalmol. 153, 789–803.e2 (2012). Taylor, H. R. Glaucoma: where to now? Ophthalmology 116, 821–822 (2009). Rein, D. B. et al. The cost-effectiveness of routine office-based identification and subsequent medical treatment of primary open-angle glaucoma in the United States. Ophthalmology 116, 823–832 (2009). Iordanous, Y., Kent, J. S., Hutnik, C. M. & Malvankar-Mehta, M. S. Projected Cost Comparison of Trabectome, iStent, and Endoscopic Cyclophotocoagulation Versus Glaucoma Medication in the Ontario Health Insurance Plan. J. Glaucoma (2013). doi:10.1097/IJG.0b013e31829d9bc7 Brusini, P. & Filacorda, S. Enhanced Glaucoma Staging System (GSS 2) for classifying functional damage in glaucoma. J. Glaucoma 15, 40–46 (2006). Mills, R. P. et al. Categorizing the stage of glaucoma from pre-diagnosis to end-stage disease. Am. J. Ophthalmol. 141, 24–30 (2006). Advanced Glaucoma Intervention Study. 2. Visual field test scoring and reliability. Ophthalmology 101, 1445–1455 (1994). Zhang, X. et al. Predicting Development of Glaucomatous Visual Field Conversion Using Baseline Fourier-Domain Optical Coherence Tomography. Am. J. Ophthalmol. 163, 29–37 (2016)."
}
]
},
{
"id": "13591",
"date": "13 May 2016",
"name": "William G. Myers",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe concept of creating glaucoma index buckets has been validated by the authors in their PLoS One paper1. They showed that similar patients with and without cataract undergoing ab-interno trabeculectomy can be combined in one study. I found the conclusion that use of the Trabectome would be potentially useful in many cases of procedures combined with cataract surgery, even when the glaucoma is fairly far advanced. As the site of greatest resistance to fluid exit, removing the trabecular membrane alone could potentially lower a high enough pressure gradient, at least temporarily. With the potential to avoid a full thickness filter and its attendant complications, the Trabectome could buy time to ride out a post operative cataract pressure spike.I would like to see more analysis of when an ab-interno trabeculectomy might be risky such as in severe glaucoma with split fixation where the alternative is an ab-externo filter or tube and then later cataract surgery. If this procedure is more effective than iStent and other bypass procedures, and involves no costly per case costs in comparison, then Trabectome with phacoemulsification may be the most prudent choice.. At the other end of the spectrum in ocular hypertension patients with no current measurable damage, is the risk profile of Trabectome combined with phacoemulsification reasonable, compared to the risk and cost of drops or SLT as alternatives? We know that the effectiveness of only one iStent, the FDA approved norm in the US, is quite often not sufficient to reduce pressures adequately.There was another issue to be considered. The grading system uses pressure as one of the components and also uses the pressure drop as the metric of success. There be some circular reasoning involved, as the initial pressure head itself is a determinant of the resulting pressure drop using drops, lasers or surgery. Unfortunately the study can not be replicated exactly due to patient confidentiality considerations.The validated glaucoma severity scale. from their previous paper in PLoS, uses simple-to-obtain inputs: IOP, number of medications and visual field loss; graded using MD and PSD from the visual field testing. The combination of these factors with definable numeric values can be utilized by cataract surgeons without glaucoma subspecialties.",
"responses": [
{
"c_id": "2017",
"date": "27 Jun 2016",
"name": "Nils Loewen",
"role": "Author Response",
"response": "I would like to see more analysis of when an ab-interno trabeculectomy might be risky such as in severe glaucoma with split fixation where the alternative is an ab-externo filter or tube and then later cataract surgery. If this procedure is more effective than iStent and other bypass procedures and involves no costly per case costs in comparison, then Trabectome with phacoemulsification may be the most prudent choice. Authors: This study is designed to analyze how patients with more severe glaucoma fared during the one-year follow-up, and the all-inclusive design maximizes the patient numbers. The results indicate that this microincisional surgery is a viable first option even in more severe glaucoma for many patients, but the resulting pressure is not low enough for many as the survival curve demonstrates. Postoperative complications are surprisingly common after traditional glaucoma surgery. Leakage, endophthalmitis, hypotony, hardware erosion, damage to ocular tissues and other complications that are vision threatening occur with an additive probability of 77% in trabeculectomies and 58% in glaucoma drainage implants.9 The surgeon has to discuss with each patient whether the avoidance of postoperative complications of traditional surgeries is worth the risk of an insufficient pressure reduction and the need to move on. In this context, it is important to recall that the postoperative IOP for tubes and trabeculectomies is only approximately 1 to 2 mmHg less that reported here.10 In forthcoming publications with fewer patients and full data disclosure, we report propensity score matched results of tube shunts and trabectome for a valid comparison of these surgical modalities. Because of the speed and safety of ab interno trabeculectomy, there is also a role for combining the two to achieve a lower IOP with fewer medications. At the other end of the spectrum in ocular hypertension patients with no current measurable damage, is the risk profile of Trabectome combined with phacoemulsification reasonable, compared to the risk and cost of drops or SLT as alternatives? We know that the effectiveness of only one iStent, the FDA approved norm in the US, is quite often not sufficient to reduce pressures adequately. Authors: It is hard to compare the effectiveness of different surgeries and in various practices without a randomized controlled design. The only way to obtain a high-quality comparison is to perform matching as we have done recently.3,5 A cost analysis is similarly specific to the geographical area and practice. Prior cost analyses suggest that SLT is better or similar to eye drops in lowering IOP and more cost effective, even returning more than initially invested.11,12 The same can be said about micro-incisional glaucoma surgeries13 although this depends on the long-term viability of the chosen surgical modality and costs of anesthesia. SLT is our first line of treatment before use of drops and trabectome surgery presents a sharp step up. It is certainly convenient, fast and safe to combine at the time of cataract surgery. After trabectome surgery, only the distal outflow tract generates pressure causing most patients to drop into the mid-teens while SLT achieves more of a relative reduction that depends on the preoperative IOP. Our study presented here shows that the postoperative results are mostly independent of preoperative IOP (or that a larger IOP reduction is achieved in higher preoperative IOP). There was another issue to be considered. The grading system uses pressure as one of the components and also use s the pressure drop as the metric of success. There be some circular reasoning involved, as the initial pressure head itself is a determinant of the resulting pressure drop using drops, lasers or surgery. Authors: Different from other glaucoma staging systems graded by a defect of visual field or optic nerve injury14–16, the glaucoma index here includes the relative resistance to treatment by individual patients. There is no circular logic, however, because we did not develop a prediction model that would have used a set of normal patients to train an algorithm and control patients within a study that were separated out from progressing patients as we have used elsewhere17. Both IOP and medications are reduced (or one could say the glaucoma index is reduced) by this procedure. Unfortunately the study can not be replicated exactly due to patient confidentiality considerations. Authors: HIPAA concerns are particularly relevant for this study describing many individuals. Fines can quickly climb to millions of dollars: https://www.truevault.com/blog/what-is-the-penalty-for-a-hipaa-violation.html We are very impressed with F1000Research and the compelling benefits of pre-printing and open peer review and will provide full open data with our upcoming publications. Although it is the standard not to release full datasets in clinical ophthalmology publications, we have come to realize that this is a disservice. For instance, one could apply matching strategies to other datasets to compare surgeries and we hope to make ours available for this purpose. The validated glaucoma severity scale from their previous paper in PLoS, uses simple-to-obtain inputs: IOP, the number of medications and visual field loss graded using MD and PSD from the visual field testing. The combination of these factors with definable numeric values can be utilized by cataract surgeons without glaucoma subspecialties. Authors: Yes, we agree. We were surprised that this has not been attempted before. Hopefully, better indices will become available. We have a glaucoma risk calculator in our medical record system, a customized version of Epic, and that has been a big step forward. References used in this Reply to Reviewers Loewen, R. T. et al. Impact of a Glaucoma Severity Index on Results of Trabectome Surgery: Larger Pressure Reduction in More Severe Glaucoma. PLoS One 11, e0151926 (2016). Bussel, I. I., Kaplowitz, K., Schuman, J. S., Loewen, N. A. & Trabectome Study Group. Outcomes of ab interno trabeculectomy with the trabectome by degree of angle opening. Br. J. Ophthalmol. 99, 914–919 (2015). Parikh, H. A., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One 11, e0149384 (2016). Wollstein, G. et al. Retinal nerve fibre layer and visual function loss in glaucoma: the tipping point. Br. J. Ophthalmol. 96, 47–52 (2012). Neiweem, A. E., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Glaucoma Surgery Calculator: Limited Additive Effect of Phacoemulsification on Intraocular Pressure in Ab Interno Trabeculectomy. PLoS One 11, e0153585 (2016). Iacus, S. M., King, G. & Porro, G. Causal Inference without Balance Checking: Coarsened Exact Matching. Polit. Anal. 20, 1–24 (2012). Parikh, H. A., Bussel, I. I., Schuman, J. S., Brown, E. N. & Loewen, N. A. Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One 11, e0149384 (2016). Akil, H. et al. Clinical Results of Ab Interno Trabeculotomy Using the Trabectome in Patients with Pigmentary Glaucoma compared to Primary Open Angle Glaucoma. Clin. Experiment. Ophthalmol. (2016). doi:10.1111/ceo.12737 Gedde, S. J. et al. in Am J Ophthalmol 153, 804–814.e1 (2012 Elsevier Inc, 2012). Gedde, S. J. et al. Treatment Outcomes in the Tube Versus Trabeculectomy (TVT) Study After Five Years of Follow-up. Am. J. Ophthalmol. 153, 789–803.e2 (2012). Taylor, H. R. Glaucoma: where to now? Ophthalmology 116, 821–822 (2009). Rein, D. B. et al. The cost-effectiveness of routine office-based identification and subsequent medical treatment of primary open-angle glaucoma in the United States. Ophthalmology 116, 823–832 (2009). Iordanous, Y., Kent, J. S., Hutnik, C. M. & Malvankar-Mehta, M. S. Projected Cost Comparison of Trabectome, iStent, and Endoscopic Cyclophotocoagulation Versus Glaucoma Medication in the Ontario Health Insurance Plan. J. Glaucoma (2013). doi:10.1097/IJG.0b013e31829d9bc7 Brusini, P. & Filacorda, S. Enhanced Glaucoma Staging System (GSS 2) for classifying functional damage in glaucoma. J. Glaucoma 15, 40–46 (2006). Mills, R. P. et al. Categorizing the stage of glaucoma from pre-diagnosis to end-stage disease. Am. J. Ophthalmol. 141, 24–30 (2006). Advanced Glaucoma Intervention Study. 2. Visual field test scoring and reliability. Ophthalmology 101, 1445–1455 (1994). Zhang, X. et al. Predicting Development of Glaucomatous Visual Field Conversion Using Baseline Fourier-Domain Optical Coherence Tomography. Am. J. Ophthalmol. 163, 29–37 (2016)."
}
]
},
{
"id": "13594",
"date": "20 May 2016",
"name": "Steven R. Sarkisian Jr.",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting study demonstrating that ab interno trabeculectomy is an appropriate surgery to attempt to control more than mild glaucoma. Ab interno trabeculotomy is a growing trend in glaucoma surgery, not just with the trabectome device, but also with the dual blade that is recently on the market from New World Medical, and also the Sight Sciences device called TRAB 360. The trend toward ab interno, non-penetrating surgery in patients with not only mild, but even advanced glaucoma demonstrates our aversion for conventional filtration surgery and the significantly high complication rate seen with it. I am grateful to the authors for writing up their large case series of results that mirror a \"real world\" practice using this procedure. Glaucoma surgeons need to consider expanding the use of this type of technique, knowing that by doing so, they do not rule out the possibility of doing conventional surgery in the future, if necessary.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-762
|
https://f1000research.com/articles/5-1512/v1
|
27 Jun 16
|
{
"type": "Review",
"title": "Advances in genetics: widening our understanding of prostate cancer",
"authors": [
"Angela C. Pine",
"Flavia F. Fioretti",
"Greg N. Brooke",
"Charlotte L. Bevan",
"Angela C. Pine",
"Flavia F. Fioretti"
],
"abstract": "Prostate cancer is a leading cause of cancer-related death in Western men. Our understanding of the genetic alterations associated with disease predisposition, development, progression, and therapy response is rapidly improving, at least in part, owing to the development of next-generation sequencing technologies. Large advances have been made in our understanding of the genetics of prostate cancer through the application of whole-exome sequencing, and this review summarises recent advances in this field and discusses how exome sequencing could be used clinically to promote personalised medicine for prostate cancer patients.",
"keywords": [
"Prostate Cancer",
"Whole Exome Sequencing",
"WES",
"Personalised Medicine",
"Androgen",
"Transcriptome",
"Metastasis",
"Cancer progression",
"DNA damage repair"
],
"content": "Introduction\n\nProstate cancer (PCa) is the most common cancer among men in the UK, with over 40,000 cases diagnosed every year1. More than 10,000 men die from the disease in the UK per annum, making it the second most common cause of cancer-related death behind only lung cancer. Similarly, in the USA, PCa accounts for just over a quarter of all cancer diagnoses in men, with 220,800 diagnoses and 27,540 deaths from PCa predicted in 20152. Statistically significant risk factors associated with PCa include ethnicity, family history of the disease, and age3, with over 75% of all PCa cases diagnosed in men over 65 years of age1. Other factors such as cigarette smoking history, lower physical activity, higher body mass index, and height are associated with increased risk of fatal disease3,4.\n\nPCa growth is dependent upon the androgen receptor (AR), a ligand-dependent transcription factor. In response to androgen binding (the major AR ligand in prostate is dihydrotestosterone), the AR regulates the expression of target genes/proteins important in PCa growth (e.g. 5–7). The treatment given to patients with PCa is dependent on the grade and stage of the disease8. PCa that is contained within the prostate capsule can be removed via surgery to remove the prostate or treated using radiotherapy9. Given the potential side effects and the fact that low-grade tumours often grow slowly and may not become clinically significant, many patients, especially if they are older, tend to be monitored rather than treated (termed “active surveillance”).\n\nSince androgenic hormones drive prostate tumour growth, therapies that target the androgen signalling pathway are commonly used for tumours that have spread outside the capsule10. These initial hormonal therapies fall into two categories11,12. The first blocks the gonadal production of androgens by pituitary downregulation. This can be achieved using luteinising hormone releasing hormone (LHRH) analogues. They cause an initial spike in androgen levels; however, within 2 weeks, castrate levels of circulating testosterone are achieved due to hyperstimulation of the hypothalamo-pituitary axis. In contrast, anti-androgen therapies (e.g. bicalutamide and enzalutamide) do not reduce androgen levels per se but act directly on the AR; their binding to it results in the AR adopting an inactive conformation with subsequent inhibition of downstream events. The two approaches may be used sequentially with a switch in treatment when the first fails, or simultaneously in complete androgen blockade. However, although such hormone therapy is initially successful in the majority of patients, it invariably eventually fails, with tumours becoming unresponsive to therapy within 1–3 years and tumours progressing to the aggressive stage termed castration-resistant PCa (CRPC). Although in recent years several new therapies have been developed with some licensed for use in CRPC13,14, there remain few effective options and the mean survival period for patients with CRPC in 2012 was just 13.5 months15. There is therefore a great need to identify therapeutic strategies to prevent/treat CRPC but also to develop the means and biomarkers to stratify patients for optimal therapy, and the genetic information from the studies described below is a major step in this process.\n\n\nWhole-exome sequencing\n\nThe human genome consists of approximately 3 × 109 base pairs. Only around 1% of these (3 × 107 base pairs) is believed to represent coding sequence, but it is estimated that 85% of disease-causing mutations are located in these coding regions of the genome – collectively termed the exome16,17. Hence, most studies to date have concentrated on characterizing the exome, initially indirectly, through microarray expression analyses, and now owing to advances in DNA sequencing technologies by whole-exome sequencing (WES). WES, on which this review focuses, has led the way in uncovering mutations in coding regions responsible for many diseases and in practical and economic terms is, for many, more feasible than whole-genome sequencing (WGS), although it will not identify changes in non-coding regions of the genome (see later)18. Hence, availability of financial resources as well as project-specific requirements (such as the ability to detect splice variants, gene fusions, and non-coding transcripts) are likely to influence decisions on whether to employ WGS, transcriptomic sequencing, WES, or other forms of targeted re-sequencing (such as deep sequencing of targeted gene panels).\n\nIn WES, DNA sequences are isolated only from exons, and data analysis compares the patient sequence to that of a reference exome aligning all of the captured exons. The variants found are compared to a control population database containing non-disease-causing variants. After the common variants are filtered out, the data can be compared to the exomes of unaffected individuals or normal tissue to identify disease-associated variants19. The first exome sequencing study to be reported was performed by Ng et al.20, in which the exomes of 12 individuals with Freeman–Sheldon syndrome (FSS), a rare dominantly inherited disorder, were sequenced. In agreement with previous reports21, the study demonstrated that mutations of embryonic myosin heavy chain (MYH3) were present in patients with the disease. Since this study, the amount of literature published using WES technology, and the range of diseases, has been increasing exponentially. Furthermore, the successful diagnostic rate of rare disease using this technology has now reached 25%22, with many clinical laboratories offering WES as a cost-effective means of clinical testing and diagnosis23.\n\nWES is preferably performed on DNA obtained from fresh or frozen patient samples. The issue of obtaining fresh PCa tissue for genome analysis was recently addressed by Menon and colleagues, who used WES to compare fresh samples and formalin-fixed paraffin-embedded (FFPE) material from the same patient24. The study demonstrated a high degree of overlap in single nucleotide variations in both types of samples, suggesting that FFPE material is a viable option for such studies. This supports previous work by Schweiger et al. and Kerick et al.25,26 and is a major consideration for PCa research, since it supports the use of archival FFPE biopsy samples for WES.\n\nThe availability of samples from advanced metastatic PCa is historically limited to biopsies taken from the primary tumour; analysis of tumours to, for example, identify mechanisms of therapy resistance has therefore been hampered by lack of material. This situation is improving, with metastatic samples taken post-mortem in systematic approaches such as the expanding “warm autopsy” program developed at the University of Michigan by Rubin, Pienta, and colleagues27 and recently the move to biopsy metastatic tumours from living patients, as exemplified in the landmark paper from Robinson et al. using both WES and transcriptomic sequencing to characterize genetic lesions in 150 metastatic CRPC patients, both at bone and soft tissue metastatic sites28. Also, in recent years, there has been a move towards non-invasive sampling (liquid biopsies such as plasma, serum, urine, and semen) to obtain samples for WES analysis and biomarkers in general. For example, Mutaza and colleagues performed WES on circulating cell-free tumour DNA (ccfDNA) obtained from the plasma of patients29. The study identified a number of mutations associated with drug resistance, e.g. an activating mutation in phosphatidylinositol-4,5-bisphosphate 3-kinase (PIK3CA) found following paclitaxel treatment. The ability to use liquid biopsies in the prostate field will circumvent issues of material availability and is an important development in the field of WES use for personalised medicine.\n\n\nIdentification of gene alterations associated with prostate cancer susceptibility\n\nSince familial PCa was first described in the 1950s30, there has been a strong search for hereditary mutations linked with the occurrence of the disease. For example, genome-wide association studies (GWAS) and single nucleotide polymorphism (SNP) arrays have identified more than 100 PCa susceptibility loci31,32. The majority of these SNPs are located in non-coding regions, and bioinformatic approaches have been used to identify candidate genes affected by these variants33.\n\nWES has also been used to identify genetic variants in coding regions that correlate with PCa predisposition. The G84E mutation in the homeobox transcription factor HOXB13 is an example of such a SNP strongly associated with early onset familial PCa, confirmed through parallel targeted sequencing of germline DNA from 94 unrelated PCa patients and their families34–36. Similarly, both BRCA1 and BRCA2 have been linked to PCa predisposition, although no specific SNP has been identified, rather a variety of genetic alterations (e.g. protein-truncating mutations, in-frame deletions, and missense variants) in the two genes that cause loss of protein function37,38. The IMPACT study showed, in fact, that targeting prostate-specific antigen (PSA) screening to men bearing BRCA mutations identified a higher proportion with PCa and that BRCA mutation-positive men are more likely to have an aggressive form of the disease39.\n\nRand et al.40 compared the exomes of 2165 PCa cases and 2034 controls of African ancestry with the aim of identifying protein-coding variations that affect disease risk in this population. Among the significant associations identified were mutations in Secreted Protein Acidic and Rich in Cysteine-Like 1 (SPARCL1) and Protein Tyrosine Phosphatase, Receptor Type, R (PTPRR). SPARCL1 has been shown to have tumour suppressor activity41, and the alanine to aspartic acid substitution at amino acid 49 is associated with reduced PCa risk (odds ratio [OR] = 0.78, p = 1.8 × 10-6). In contrast, the substitution identified in PTPRR (Val239Ile), an AR target gene and regulator of the RAS/ERK1/2 pathway42, was associated with increased risk (OR = 1.62, p = 2.5 × 10-5). In another WES study of individuals from families with three or more affected individuals, it is notable that several of the changes associated with PCa were in genes linked to DNA damage repair, including three poly(adenosine diphosphate [ADP]-ribose) polymerase (PARP) genes43; this is especially interesting given recent reports of clinical benefit conferred by PARP inhibitors in PCa patients with defects in DNA-repair genes44.\n\n\nGene alterations associated with prostate cancer development/progression\n\nA number of genetic alterations have been correlated with PCa development and progression. Perhaps unsurprisingly, the recent comprehensive profiling study of metastatic CRPC using WES and transcriptomic sequencing found mutations in genes in the AR signalling pathway in over 71% of cases: the majority were in the AR gene itself28. Since Taplin et al.’s original report of AR mutation and Visakorpi et al.’s report of AR amplification in advanced PCa, many more have been published45,46. These alterations are rare in early stages of the disease and appear in response to selective pressure resulting from the hormone therapies administered, allowing the receptor to continue to drive growth in CRPC10,47–53. There are many reports of missense mutations leading to amino acid substitutions, usually within the ligand-binding region, which broaden ligand specificity to allow activation by, for example, adrenal androgens, glucocorticoids, and even therapeutic antiandrogens54–58. More recently, constitutively active AR splice variants have been identified, which circumvent the requirement for ligand59,60. To date, at least 20 such variants have been described, all of which lack the ligand-binding domain, are nuclear in the absence of ligand, and have been reported to have constitutive ligand-independent activity, although some studies suggest they still require the presence of the full-length receptor for activity60–64. As well as alterations of AR, mutations of other components of the AR signalling pathway have been found to correlate with disease progression. In their landmark study, Grasso et al. compared the exomes of 50 heavily treated metastatic CRPC tumours with 11 high-grade treatment-naïve non-metastatic tumours65, and a number of genes encoding proteins that interact with, and/or regulate the activity of, the AR were found to be altered. For example, mixed-lineage leukemia protein 2/histone-lysine N-methyltransferase 2D (MLL2/KMT2D), a histone-modifying enzyme that interacts with the AR, was mutated in 8.6% of cancers. Alterations were also found in FOXA1 (3.4% of cases), a pioneer factor that de-compacts DNA allowing genomic access of nuclear receptors, including the AR66. The majority of FOXA1 mutations and indels identified were in the carboxy-terminal transactivation domain, and functional assays demonstrated that these alterations enhanced tumour growth. In support of other studies, AR mutations and an increase in copy number were also identified in the majority of patients with advanced disease65. In addition, the recent Robinson et al. paper highlighted that 71.3% of metastatic CRPC tumours carried AR pathway mutations, the majority in the AR itself but others in e.g. AR cofactors (NCoR1/2) and, again, the pioneer factor FOXA128.\n\nThe gene often cited as most frequently mutated in primary PCa, encoding Speckle-type POZ protein (SPOP), is mutated in over 10% of primary prostate tumours67. Mutations in this gene appear early in development and impact the ability of the cell to repair DNA damage, leading to genomic instability68. Other targets of SPOP include the AR and the ERG oncogene, confirming the importance of this gene in driving tumour progression69.\n\nPhosphatase and Tensin homolog (PTEN) is found mutated at a similar frequency in primary PCa (10%), while PTEN deletion occurs in up to 70% of surgically treated cancers and over 60% of metastatic prostate tumours70–73. PTEN mutations or copy loss leads to increased PI3K/Akt signalling, which translates into cell survival and proliferation, e.g. through ligand-independent activation of the AR signaling pathway74. It has been hypothesised that PTEN deletion creates genomic instability that then facilitates other alterations, e.g. the TMPRSS-ERG fusion commonly found in PCa75,76. The TMPRSS-ERG gene originates from fusion between the TMPRSS2 promoter, which is androgen responsive, and the ERG gene77. ERG is a member of the ETS family of transcription factors, which has roles in numerous processes including cell proliferation, apoptosis, differentiation, angiogenesis, and invasiveness. This gene fusion causes the oncogene ERG to be under the control of the androgen inducible TMPRSS2 promoter, which appears to have a subsequent bearing upon tumour progression78. Although ERG is the most common fusion partner, other ETS genes (notably ETV1 and ETV5) can be fused to the TMPRS22 promoter in prostate tumours, and also mutations in ETS gene family members have been identified in tumours, prompting speculation that some of these may have tumour suppressive function65,79.\n\nLoss of function of tumour suppressors is also a common event in PCa development and progression, and those frequently lost are p53, retinoblastoma (Rb), and NK3 transcription factor related, locus 1 (NKX3.1)80–82. Dysregulation of the tumour suppressor p53 is rarer in PCa than other tumour types, at around 5–10% in primary tumours, but increases to around 50% in metastatic CRPC28,67,83. A recent study, using candidate gene exome sequencing, suggested that patients with dominant negative p53 mutations had the worst outcome, and this feature by itself has independent prognostic relevance for patients84. NKX3.1 is an androgen-regulated gene known to regulate prostate organogenesis during embryogenesis and is expressed throughout adult life, where it regulates ductal function and secretion85. Complete loss of NKX3.1 expression is evident in 5% of benign prostatic hyperplasias, 20% of high-grade prostatic intraepithelial neoplasias, 34% of hormone-refractory PCa, and 78% of metastases, supporting a role in disease progression65,82. In contrast to p53 and Rb, NKX3.1 tumour suppressor activity is restricted to the prostate and its loss of activity is usually due to absent protein expression rather than inactivating mutations86.\n\nWhen comparing WES studies, some common alterations are evident. Taking the top 30 hits each from two recent comparable exome studies of advanced metastatic PCa, 17 genes were found to be altered in both studies28,65 (Figure 1). For example, alterations in AR, TP53, ETS fusion, PTEN, and RB1 were common to both studies. However, a number of other alterations were found in only one study, some with a relatively high incidence rate (e.g. Grasso et al. found CYP11B1 to be altered in 20% of patients65). The discrepancies between studies are likely to be resolved as the number of tumours sequenced increases but are also likely to represent the significant heterogeneity associated with PCa. Inevitably, key driver mutations will be found in multiple studies. Further comparison with other larger sequencing studies such as the 100,000 human genome project87 will aid in the identification of the key/driver mutations and variants important in PCa initiation or progression, as these will be enriched in PCa compared to other diseases and the general population.\n\nThe top 30 genetic alterations found in each of the studies by Robinson et al. and Grasso et al.28,65 were compared.\n\n\nThe use of whole-exome sequencing for personalised medicine\n\nThere is an urgent need to stratify patients according to which therapeutic is likely to be most effective, to increase drug efficacy, and to reduce over-treatment and unnecessary side effects. WES holds great promise in this regard and has already been demonstrated to be a useful tool in terms of determining the cause of resistance to therapeutics or indeed why certain unconventional treatments may benefit patients with a particular cancer for which they would not normally be given. An example of this was an unexpected finding in a study conducted by Beltran and colleagues88, which used WES to analyse the exomes of 97 patients with a range of metastatic cancers. One of the tumours analysed was from a PCa patient found to have an exceptional response to cisplatin treatment. Exome sequencing identified that the DNA repair protein FANCA had reduced expression and activity as a result of somatic hemizygous deletion and a partial loss of function as a result of a germline missense mutation in the second allele; subsequent assays demonstrated that loss of FANCA function was associated with platinum hypersensitivity, thus providing a rationale for the patient’s clinical response to an unconventional treatment. The widespread application of prospective WES in personalised medicine was also demonstrated in this study, since the authors were able to identify therapeutics (approved or in development), for 94% of the patients, expected to be effective given the exome profile generated88. Robinson et al. reported a similar rate of “actionable” mutations in their study of metastatic CRPC, i.e. mutations on the basis of which informed treatment advice could be offered, including BRCA or ATM mutations that indicate use of PARP inhibitors28. To date, each exome or transcriptome sequencing study of the prostate identifies a number of mutations unique to that study, while also a notable number of common mutations and/or affected genes. This provides both evidence for the key pathways and driver mutations in PCa and potential information leading to the effective application of a wide range of therapeutics.\n\n\nConcluding remarks\n\nThe launch of the International Cancer Genome Consortium89 (https://icgc.org) in 2008 paved the way for genome studies on over 50 cancer types and through the use of sequencing approaches, including WES, has significantly improved our understanding of the genomic, transcriptomic, and epigenomic changes associated with different tumour types. Repositories such as this are providing a valuable resource for researchers in the field. In comparison to complete genome sequencing, WES provides information only about alterations in the coding sequence; however, it is cost effective and the reduced data analysis associated with WES means that it is likely to continue to be a valuable tool for PCa research. It also holds great promise in the clinic, having the potential to assist and inform personalised medicine for men with the disease. Undoubtedly in the future, as transcriptomic approaches become more widely used, they will lead to similar advances relating to the non-coding portion of the genome; microRNAs, long non-coding RNAs, other non-coding RNAs, and epigenetic changes are also likely to yield markers for tumour classification as well as actionable mutations. Despite the seemingly unlimited potential of WES, the frequent lack of knowledge of the functional consequences of such gene alterations is an issue that requires addressing in subsequent functional studies. These will be invaluable in characterising the phenotypic consequences of these gene alterations and are likely to yield findings that can be exploited for therapeutic gain.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThe authors are grateful for support from Prostate Cancer UK (S12-026) and Cancer Research UK (C42671/A12990) during the writing of this review.\n\n\nReferences\n\nCRUK: Cancer mortality for common cancers. 2014. [cited 2014 6/11/14]. Reference Source\n\nSiegel RL, Miller KD, Jemal A: Cancer statistics, 2015. CA Cancer J Clin. 2015; 65(1): 5–29. PubMed Abstract | Publisher Full Text\n\nGiovannucci E, Liu Y, Platz EA, et al.: Risk factors for prostate cancer incidence and progression in the health professionals follow-up study. Int J Cancer. 2007; 121(7): 1571–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMöller E, Wilson KM, Batista JL, et al.: Body size across the life course and prostate cancer in the Health Professionals Follow-up Study. Int J Cancer. 2016; 138(4): 853–65. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBrooke GN, Gamble SC, Hough MA, et al.: Antiandrogens act as selective androgen receptor modulators at the proteome level in prostate cancer cells. Mol Cell Proteomics. 2015; 14(5): 1201–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrooke GN, Culley RL, Dart DA, et al.: FUS/TLS is a novel mediator of androgen-dependent cell-cycle progression and prostate cancer growth. Cancer Res. 2011; 71(3): 914–24. PubMed Abstract | Publisher Full Text\n\nGamble SC, Chotai D, Odontiadis M, et al.: Prohibitin, a protein downregulated by androgens, represses androgen receptor activity. Oncogene. 2007; 26(12): 1757–68. PubMed Abstract | Publisher Full Text\n\nTrewartha D, Carter K: Advances in prostate cancer treatment. Nat Rev Drug Discov. 2013; 12: 823–4. PubMed Abstract | Publisher Full Text\n\nCRUK: About hormone therapy for prostate cancer. 2014. Reference Source\n\nBrooke GN, Bevan CL: The role of androgen receptor mutations in prostate cancer progression. Curr Genomics. 2009; 10(1): 18–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoktas S, Crawford ED: Optimal hormonal therapy for advanced prostatic carcinoma. Semin Oncol. 1999; 26(2): 162–73. PubMed Abstract\n\nHeidenreich A, Bastian PJ, Bellmunt J, et al.: EAU guidelines on prostate cancer. Part II: Treatment of advanced, relapsing, and castration-resistant prostate cancer. Eur Urol. 2014; 65(2): 467–79. PubMed Abstract | Publisher Full Text\n\nCrawford ED, Higano CS, Shore ND, et al.: Treating Patients with Metastatic Castration Resistant Prostate Cancer: A Comprehensive Review of Available Therapies. J Urol. 2015; 194(6): 1537–47. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nFerraldeschi R, Welti J, Luo J, et al.: Targeting the androgen receptor pathway in castration-resistant prostate cancer: progresses and prospects. Oncogene. 2015; 34(14): 1745–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHirst CJ, Cabrera C, Kirby M: Epidemiology of castration resistant prostate cancer: a longitudinal analysis using a UK primary care database. Cancer Epidemiol. 2012; 36(6): e349–53. PubMed Abstract | Publisher Full Text\n\nBotstein D, Risch N: Discovering genotypes underlying human phenotypes: past successes for mendelian disease, future approaches for complex disease. Nat Genet. 2003; 33Suppl: 228–37. PubMed Abstract | Publisher Full Text\n\nMajewski J, Schwartzentruber J, Lalonde E, et al.: What can exome sequencing do for you? J Med Genet. 2011; 48(9): 580–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRabbani B, Tekin M, Mahdieh N: The promise of whole-exome sequencing in medical genetics. J Hum Genet. 2014; 59(1): 5–15. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nProws CA, Tran G, Blosser B: Whole exome or genome sequencing: nurses need to prepare families for the possibilities. J Adv Nurs. 2014; 70(12): 2736–45. PubMed Abstract | Publisher Full Text\n\nNg SB, Turner EH, Robertson PD, et al.: Targeted capture and massively parallel sequencing of 12 human exomes. Nature. 2009; 461(7261): 272–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nToydemir RM, Rutherford A, Whitby FG, et al.: Mutations in embryonic myosin heavy chain (MYH3) cause Freeman-Sheldon syndrome and Sheldon-Hall syndrome. Nat Genet. 2006; 38(5): 561–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSeaby EG, Pengelly RJ, Ennis S: Exome sequencing explained: a practical guide to its clinical application. Brief Funct Genomics. 2015; pii: elv054. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWilliams ES, Hegde M: Implementing genomic medicine in pathology. Adv Anat Pathol. 2013; 20(4): 238–44. PubMed Abstract | Publisher Full Text\n\nMenon R, Deng M, Rüenauver K, et al.: Somatic copy number alterations by whole-exome sequencing implicates YWHAZ and PTK2 in castration-resistant prostate cancer. J Pathol. 2013; 231(4): 505–16. PubMed Abstract | Publisher Full Text\n\nSchweiger MR, Kerick M, Timmermann B, et al.: Genome-wide massively parallel sequencing of formaldehyde fixed-paraffin embedded (FFPE) tumor tissues for copy-number- and mutation-analysis. PLoS One. 2009; 4(5): e5548. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKerick M, Isau M, Timmermann B, et al.: Targeted high throughput sequencing in clinical cancer settings: formaldehyde fixed-paraffin embedded (FFPE) tumor tissues, input amount and tumor heterogeneity. BMC Med Genomics. 2011; 4: 68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRubin MA, Putzi M, Mucci N, et al.: Rapid (\"warm\") autopsy study for procurement of metastatic prostate cancer. Clin Cancer Res. 2000; 6(3): 1038–45. PubMed Abstract\n\nRobinson D, van Allen EM, Wu Y, et al.: Integrative clinical genomics of advanced prostate cancer. Cell. 2015; 161(5): 1215–28. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMurtaza M, Dawson S, Tsui DW, et al.: Non-invasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature. 2013; 497(7447): 108–12. PubMed Abstract | Publisher Full Text\n\nGianferrari L, Arrigoni G, Cresseri A, et al.: [Genetic and clinico-statistical research on neoplasms of the prostate]. Acta Genet Med Gemellol (Roma). 1956; 5(2): 224–33. PubMed Abstract\n\nGoh CL, Schumacher FR, Easton D, et al.: Genetic variants associated with predisposition to prostate cancer and potential clinical implications. J Intern Med. 2012; 271(4): 353–65. PubMed Abstract | Publisher Full Text\n\nEeles RA, Olama AA, Benlloch S, et al.: Identification of 23 new prostate cancer susceptibility loci using the iCOGS custom genotyping array. Nat Genet. 2013; 45(4): 385–91, 391e1–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThibodeau SN, French AJ, McDonnell SK, et al.: Identification of candidate genes for prostate cancer-risk SNPs utilizing a normal prostate tissue eQTL data set. Nat Commun. 2015; 6: 8653. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nEwing CM, Ray AM, Lange EM, et al.: Germline mutations in HOXB13 and prostate-cancer risk. N Engl J Med. 2012; 366(2): 141–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBreyer JP, Avritt TG, McReynolds KM, et al.: Confirmation of the HOXB13 G84E germline mutation in familial prostate cancer. Cancer Epidemiol Biomarkers Prev. 2012; 21(8): 1348–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkbari MR, Trachtenberg J, Lee J, et al.: Association between germline HOXB13 G84E mutation and risk of prostate cancer. J Natl Cancer Inst. 2012; 104(16): 1260–2. PubMed Abstract | Publisher Full Text\n\nKote-Jarai Z, Leongamornlert D, Saunders E, et al.: BRCA2 is a moderate penetrance gene contributing to young-onset prostate cancer: implications for genetic testing in prostate cancer patients. Br J Cancer. 2011; 105(8): 1230–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaia S, Cardoso M, Paulo P, et al.: The role of germline mutations in the BRCA1/2 and mismatch repair genes in men ascertained for early-onset and/or familial prostate cancer. Fam Cancer. 2016; 15(1): 111–21. PubMed Abstract | Publisher Full Text\n\nBancroft EK, Page EC, Castro E, et al.: Targeted prostate cancer screening in BRCA1 and BRCA2 mutation carriers: results from the initial screening round of the IMPACT study. Eur Urol. 2014; 66(3): 489–99. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRand KA, Rohland N, Tandon A, et al.: Whole-exome sequencing of over 4100 men of African ancestry and prostate cancer risk. Hum Mol Genet. 2016; 25(2): 371–81. PubMed Abstract | Publisher Full Text\n\nXiang Y, Qiu Q, Jiang M, et al.: SPARCL1 suppresses metastasis in prostate cancer. Mol Oncol. 2013; 7(6): 1019–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Lafferty NP, Kalna G, et al.: Androgen-regulation of the protein tyrosine phosphatase PTPRR activates ERK1/2 signalling in prostate cancer cells. BMC Cancer. 2015; 15: 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnson AM, Zuhlke KA, Plotts C, et al.: Mutational landscape of candidate genes in familial prostate cancer. Prostate. 2014; 74(14): 1371–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMateo J, Carreira S, Sandhu S, et al.: DNA-Repair Defects and Olaparib in Metastatic Prostate Cancer. N Engl J Med. 2015; 373(18): 1697–708. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTaplin ME, Bubley GJ, Shuster TD, et al.: Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med. 1995; 332(21): 1393–8. PubMed Abstract | Publisher Full Text\n\nVisakorpi T, Hyytinen E, Koivisto P, et al.: In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet. 1995; 9(4): 401–6. PubMed Abstract | Publisher Full Text\n\nTaplin ME, Rajeshkumar B, Halabi S, et al.: Androgen receptor mutations in androgen-independent prostate cancer: Cancer and Leukemia Group B Study 9663. J Clin Oncol. 2003; 21(14): 2673–8. PubMed Abstract | Publisher Full Text\n\nLamb DJ, Puxeddu E, Malik N, et al.: Molecular analysis of the androgen receptor in ten prostate cancer specimens obtained before and after androgen ablation. J Androl. 2003; 24(2): 215–25. PubMed Abstract | Publisher Full Text\n\nEvans BA, Harper ME, Daniells CE, et al.: Low incidence of androgen receptor gene mutations in human prostatic tumors using single strand conformation polymorphism analysis. Prostate. 1996; 28(3): 162–71. PubMed Abstract | Publisher Full Text\n\nMarcelli M, Ittmann M, Mariani S, et al.: Androgen receptor mutations in prostate cancer. Cancer Res. 2000; 60(4): 944–9. PubMed Abstract\n\nAntonarakis ES, Lu C, Wang H, et al.: AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med. 2014; 371(11): 1028–38. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRomanel A, Gasi Tandefelt D, Conteduca V, et al.: Plasma AR and abiraterone-resistant prostate cancer. Sci Transl Med. 2015; 7(312): 312re10. PubMed Abstract | Publisher Full Text\n\nWare KE, Garcia-Blanco MA, Armstrong AJ, et al.: Biologic and clinical significance of androgen receptor variants in castration resistant prostate cancer. Endocr Relat Cancer. 2014; 21(4): T87–T103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreen SM, Mostaghel EA, Nelson PS: Androgen action and metabolism in prostate cancer. Mol Cell Endocrinol. 2012; 360(1–2): 3–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHay CW, McEwan IJ: The impact of point mutations in the human androgen receptor: classification of mutations on the basis of transcriptional activity. PLoS One. 2012; 7(3): e32514. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeldscholte J, Berrevoets CA, Ris-Stalpers C, et al.: The androgen receptor in LNCaP cells contains a mutation in the ligand binding domain which affects steroid binding characteristics and response to antiandrogens. J Steroid Biochem Mol Biol. 1992; 41(3–8): 665–9. PubMed Abstract\n\nBrooke GN, Parker MG, Bevan CL: Mechanisms of androgen receptor activation in advanced prostate cancer: differential co-activator recruitment and gene expression. Oncogene. 2008; 27(21): 2941–50. PubMed Abstract | Publisher Full Text\n\nJoseph JD, Lu N, Qian J, et al.: A clinically relevant androgen receptor mutation confers resistance to second-generation antiandrogens enzalutamide and ARN-509. Cancer Discov. 2013; 3(9): 1020–9. PubMed Abstract | Publisher Full Text\n\nChan SC, Li Y, Dehm SM: Androgen receptor splice variants activate androgen receptor target genes and support aberrant prostate cancer cell growth independent of canonical androgen receptor nuclear localization signal. J Biol Chem. 2012; 287(23): 19736–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu R, Dunn TA, Wei S, et al.: Ligand-independent androgen receptor variants derived from splicing of cryptic exons signify hormone-refractory prostate cancer. Cancer Res. 2009; 69(1): 16–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu R, Isaacs WB, Luo J: A snapshot of the expression signature of androgen receptor splicing variants and their distinctive transcriptional activities. Prostate. 2011; 71(15): 1656–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDehm SM, Schmidt LJ, Heemers HV, et al.: Splicing of a novel androgen receptor exon generates a constitutively active androgen receptor that mediates prostate cancer therapy resistance. Cancer Res. 2008; 68(13): 5469–77. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLi Y, Chan SC, Brand LJ, et al.: Androgen receptor splice variants mediate enzalutamide resistance in castration-resistant prostate cancer cell lines. Cancer Res. 2013; 73(2): 483–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWatson PA, Chen YF, Balbas MD, et al.: Constitutively active androgen receptor splice variants expressed in castration-resistant prostate cancer require full-length androgen receptor. Proc Natl Acad Sci U S A. 2010; 107(39): 16759–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrasso CS, Wu YM, Robinson DR, et al.: The mutational landscape of lethal castration-resistant prostate cancer. Nature. 2012; 487(7406): 239–43. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYang YA, Yu J: Current perspectives on FOXA1 regulation of androgen receptor signaling and prostate cancer. Genes Dis. 2015; 2(2): 144–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarbieri CE, Baca SC, Lawrence MS, et al.: Exome sequencing identifies recurrent SPOP, FOXA1 and MED12 mutations in prostate cancer. Nat Genet. 2012; 44(6): 685–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoysen G, Barbieri CE, Prandi D, et al.: SPOP mutation leads to genomic instability in prostate cancer. eLife. 2015; 4: pii: e09207. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAn J, Ren S, Murphy SJ, et al.: Truncated ERG Oncoproteins from TMPRSS2-ERG Fusions Are Resistant to SPOP-Mediated Proteasome Degradation. Mol Cell. 2015; 59(6): 904–16. PubMed Abstract | Publisher Full Text\n\nFeilotter HE, Nagai MA, Boag AH, et al.: Analysis of PTEN and the 10q23 region in primary prostate carcinomas. Oncogene. 1998; 16(13): 1743–8. PubMed Abstract | Publisher Full Text\n\nWhang YE, Wu X, Suzuki H, et al.: Inactivation of the tumor suppressor PTEN/MMAC1 in advanced human prostate cancer through loss of expression. Proc Natl Acad Sci U S A. 1998; 95(9): 5246–50. PubMed Abstract | Free Full Text\n\nWang SI, Parsons R, Ittmann M: Homozygous deletion of the PTEN tumor suppressor gene in a subset of prostate adenocarcinomas. Clin Cancer Res. 1998; 4(3): 811–5. PubMed Abstract\n\nChen Z, Trotman LC, Shaffer D, et al.: Crucial role of p53-dependent cellular senescence in suppression of Pten-deficient tumorigenesis. Nature. 2005; 436(7051): 725–30. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFeldman BJ, Feldman D: The development of androgen-independent prostate cancer. Nat Rev Cancer. 2001; 1(1): 34–45. PubMed Abstract | Publisher Full Text\n\nBismar TA, Yoshimoto M, Vollmer RT, et al.: PTEN genomic deletion is an early event associated with ERG gene rearrangements in prostate cancer. BJU Int. 2011; 107(3): 477–85. PubMed Abstract | Publisher Full Text\n\nPhin S, Moore MW, Cotter PD: Genomic Rearrangements of PTEN in Prostate Cancer. Front Oncol. 2013; 3: 240. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTomlins SA, Rhodes DR, Perner S, et al.: Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science. 2005; 310(5748): 644–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBarry Delongchamps N: Prostate cancer: review in 2014. Diagn Interv Imaging. 2014; 95(7–8): 739–42. PubMed Abstract | Publisher Full Text\n\nDemichelis F, Setlur SR, Beroukhim R, et al.: Distinct genomic aberrations associated with ERG rearranged prostate cancer. Genes Chromosomes Cancer. 2009; 48(4): 366–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBookstein R, Shew JY, Chen PL, et al.: Suppression of tumorigenicity of human prostate carcinoma cells by replacing a mutated RB gene. Science. 1990; 247(4943): 712–5. PubMed Abstract | Publisher Full Text\n\nQian J, Hirasawa K, Bostwick DG, et al.: Loss of p53 and c-myc overrepresentation in stage T2-3N1-3M0 prostate cancer are potential markers for cancer progression. Mod Pathol. 2002; 15(1): 35–44. PubMed Abstract | Publisher Full Text\n\nBowen C, Bubendorf L, Voeller HJ, et al.: Loss of NKX3.1 expression in human prostate cancers correlates with tumor progression. Cancer Res. 2000; 60(21): 6111–5. PubMed Abstract\n\nBettendorf O, Schmidt H, Staebler A, et al.: Chromosomal imbalances, loss of heterozygosity, and immunohistochemical expression of TP53, RB1, and PTEN in intraductal cancer, intraepithelial neoplasia, and invasive adenocarcinoma of the prostate. Genes Chromosomes Cancer. 2008; 47(7): 565–72. PubMed Abstract | Publisher Full Text\n\nKluth M, Harasimowicz S, Burkhardt L, et al.: Clinical significance of different types of p53 gene alteration in surgically treated prostate cancer. Int J Cancer. 2014; 135(6): 1369–80. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBhatia-Gaur R, Donjacour AA, Sciavolino PJ, et al.: Roles for Nkx3.1 in prostate development and cancer. Genes Dev. 1999; 13(8): 966–77. PubMed Abstract | Free Full Text\n\nKim MJ, Cardiff RD, Desai N, et al.: Cooperativity of Nkx3.1 and Pten loss of function in a mouse model of prostate carcinogenesis. Proc Natl Acad Sci U S A. 2002; 99(5): 2884–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiva N: UK gears up to decode 100,000 genomes from NHS patients. Lancet. 2015; 385(9965): 103–4. PubMed Abstract | Publisher Full Text\n\nBeltran H, Eng K, Mosquera JM, et al.: Whole-Exome Sequencing of Metastatic Cancer and Biomarkers of Treatment Response. JAMA Oncol. 2015; 1(4): 466–74. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nInternational Cancer Genome Consortium.2016. Reference Source"
}
|
[
{
"id": "14604",
"date": "27 Jun 2016",
"name": "Iain McEwan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14605",
"date": "27 Jun 2016",
"name": "Nigel Mongan",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1512
|
https://f1000research.com/articles/5-1511/v1
|
27 Jun 16
|
{
"type": "Review",
"title": "Thirty-five years of research into ribozymes and nucleic acid catalysis: where do we stand today?",
"authors": [
"Sabine Müller",
"Bettina Appel",
"Darko Balke",
"Robert Hieronymus",
"Claudia Nübel",
"Bettina Appel",
"Darko Balke",
"Robert Hieronymus",
"Claudia Nübel"
],
"abstract": "Since the discovery of the first catalytic RNA in 1981, the field of ribozyme research has developed from the discovery of catalytic RNA motifs in nature and the elucidation of their structures and catalytic mechanisms, into a field of engineering and design towards application in diagnostics, molecular biology and medicine. Owing to the development of powerful protocols for selection of nucleic acid catalysts with a desired functionality from random libraries, the spectrum of nucleic acid supported reactions has greatly enlarged, and importantly, ribozymes have been accompanied by DNAzymes. Current areas of research are the engineering of allosteric ribozymes for artificial regulation of gene expression, the design of ribozymes and DNAzymes for medicinal and environmental diagnostics, and the demonstration of RNA world relevant ribozyme activities. In addition, new catalytic motifs or novel genomic locations of known motifs continue to be discovered in all branches of life by the help of high-throughput bioinformatic approaches. Understanding the biological role of the catalytic RNA motifs widely distributed in diverse genetic contexts belongs to the big challenges of future RNA research.",
"keywords": [
"ribozyme",
"catalytic",
"RNA"
],
"content": "Introduction\n\nNowadays, the term 'ribozyme' to designate an RNA catalyst is used with the same implicitness as the term ‘enzyme’ has always been used for proteinaceous biocatalysts. The fact that RNA can cleave and ligate itself, that cleavage of the 5′-trailer of tRNA in tRNA processing is mediated by the RNA subunit of RNase P, that introns may undergo self-splicing, and that the spliceosome and, even more impressively, the ribosome are actually ribozymes meanwhile has found entry into the textbooks. The exciting field of research into RNA catalysis started more than 30 years ago and over the first two decades was dominated by the discovery and identification of several classes of ribozymes occurring in nature and the elucidation of their catalytic structures and mechanisms. Apart from the ribosome that catalyses the formation of the peptide bond, all ribozymes discovered so far in nature support cleavage or ligation of a phosphodiester bond or both. However, the powerful method of SELEX (systematic evolution of ligands by exponential enrichment), originally developed for the selection of high-affinity RNA binders (aptamers) from a random library1,2, was adapted to the selection of ribozymes (and, moreover, DNAzymes) to catalyse a broad range of reactions, thus greatly enhancing the spectrum of nucleic acid catalysis3,4. Over the years, an enormous amount of data were obtained on high-resolution structures and the mechanisms of ribozymes5,6. All of this contributed to an understanding of ribozyme catalysis to an extent that has allowed engineering of ribozymes and DNAzymes with pre-determined functionality. Thus, the past decade has seen impressive developments based on the usage of known catalytic motifs in ribozyme-based switches for therapeutic and environmental diagnostics7 and more recently for control of gene expression8. In parallel, the ability of RNA to catalyse a wide variety of chemical reactions has revitalised the RNA world hypothesis, a postulated period in the origin of life in which RNA was the main player, for one as carrier of genetic information and for the other as catalyst9. Early life may have started with self-replicating RNA, and a great deal of effort has been invested in developing ribozymes capable of self-replication10 or, even more challenging, of catalysing RNA polymerisation11,12. The interest in RNA-world-relevant ribozyme activities continues, and one may well expect that there will be more to come.\n\nIn addition to ribozyme-based applications and RNA world scenarios as central topics of current research in the field, the frequency of ribozymes in nature and their function are of ongoing interest. With the help of high-throughput bioinformatic approaches, new ribozymes or novel genomic locations of known catalytic RNA motifs in highly diverse genetic contexts have been discovered in all branches of life13–16, and current research addresses the question of their biological role.\n\nThe field of ribozyme research has changed the focus from discovery and mechanistic/structural characterisation of ribozymes towards functional engineering into application. Nevertheless, the excitement of the first days of ribozyme discovery has carried over throughout the years; the search for new ribozymes or just ribozyme locations continues in all kingdoms of life—in particular, in the human genome. Moreover, as mentioned above, the search for RNA-world-relevant ribozyme activities continues with unchanged curiosity. A number of excellent review articles have summarised the achievements in nucleic acid catalysis (for recent examples, see 5,17–19). Here, we will concentrate on recent discoveries and developments in the field to draw a concise picture of ribozyme research and RNA and DNA catalysis 35 years after its beginning.\n\n\nRibozyme-based switches\n\nOver the past decade, it has become increasingly clear that the conformational flexibility of RNA is an important determinant of cellular function. In this regard, riboswitches located in the 5′-untranslated region (5′-UTR) of specific mRNAs have gained much attention20. Composed of an aptamer and an expression platform, riboswitches regulate, in a ligand-dependent manner, gene expression at the level of transcription or, alternatively, translation. Binding of a specific ligand to the aptamer induces a conformational change in the expression platform, turning gene expression ON or OFF. Interestingly, this principle of allosteric regulation was used in the test tube before it was discovered in nature21. By combination of ribozymes with aptamers, ribozyme activity was rendered ligand dependent and, consequently, adjustable. RNA or DNA aptamers for binding to a desired molecule can be produced by SELEX and linked to the ribozyme via a communication module, a sequence that translates the binding event occurring in the aptamer unit into an activity-associated conformational change within the ribozyme part. Thus, ribozyme activity can be used as a readout for a binding event, which in the case of a multiple turnover reaction would even lead to signal amplification. Owing to their modular composition of DNAzyme or ribozyme and aptamer, such constructs were termed aptazymes22. Beyond the significance of aptazymes for medicinal diagnostics and therapy, RNA- and especially DNAzyme-based biosensors have gained importance as tools in environmental monitoring, in particular to detect environmental pollutants, such as toxic heavy metals, air- and water-borne microbes, and other toxins23.\n\nAllosteric regulation of ribozyme activity has been used in a variety of contexts in the life sciences. Here, significant effort has been made in artificially modulating gene expression by a chemical signal. A ribozyme-based device positioned in the 5′- or 3′-UTR of a transcript and acting as a regulatory unit is partitioned between two functional conformations: one representing a ribozyme active state, the other an inactive state24,25. Ligand binding would support one of the two states, dependent on the specific design. As a consequence of ligand binding, translation is switched ON or OFF (Figure 1). The advantage is that the effector molecule (ligand) binds directly to the regulatory module, without the involvement of proteins, such as transcription factors, which usually mediate genetic control. After some pioneering work in the early 2000s, important progress has been made in engineering ligand-dependent ribozyme modules that switch expression of suitable reporter genes, often using the hammerhead ribozyme to control stability of the target transcript26. Furthermore, the genomic hepatitis delta virus ribozyme was engineered to control gene expression in mammalian cells and, when placed in tandem configuration, to construct a NOR logic gate device, demonstrating the modular composition of ribozyme-based RNA devices27. Moreover, the recently discovered twister ribozyme, a highly flexible and active endonucleolytic ribozyme, has been used for the development of genetic switches28. In all of these approaches, stability of a target transcript is modulated through conditional control of the cleavage activity of a ribozyme conjugated with a naturally occurring or in vitro selected aptamer domain and placed at a suitable position of the transcript. Ribozyme-based genetic control has been performed in different organisms27,29,30 and in response to diverse ligands26,30–32. In addition to chemical signals (ligand-responsive switches), physical signals (light or temperature) can be used to control ribozyme activity in such devices33. Beyond regulation of bacterial or mammalian genes, the potential of ribozyme-based genetic switches for regulation of DNA and RNA viruses has been demonstrated34. In particular, genome replication, infectious particle production and cytotoxicity of adenoviruses, and (in the case of a measles virus) progeny infectivity and virus spread were reduced by aptazyme-mediated control of gene expression, paving the way for future applications in medicine and virology.\n\nRibozyme-based ON (a) and OFF (b) switches. The ribozyme-based device is positioned in the 5′-untranslated region (5′-UTR) of the transcript of interest. (a) In the absence of a specific ligand, the ribozyme is inactive and the ribosome-binding site (RBS) is sequestered in a double-stranded region; translation is switched OFF. Upon ligand binding, the ribozyme is activated and cleavage can take place. As a result, the RBS is set free and translation can proceed. (b) In the absence of a specific ligand, the ribozyme undergoes self-cleavage, thereby freeing the RBS and allowing translation to proceed. Binding of the ligand inhibits ribozyme activity, and translation is switched OFF.\n\nImportant challenges in the engineering of ribozyme-based switches by modular composition are the link between ribozyme (actuator) and ligand-responsive aptamer (sensor) and the relatively slow kinetics of secondary structure changes induced by ligand binding, thus limiting the regulatory potential of ribozyme-based switches24,35. Therefore, it is all the more important that powerful protocols for in vivo selection and screening and for high-throughput cellular RNA device engineering have been developed24–26,28,36. In general ribozyme-based switches allow for the regulation of gene expression by up to 30-fold26,27. However, it can be anticipated that, based on novel protocols for RNA device engineering and on the ever-growing understanding of the underlying structure-function relationships, novel designs will outperform those currently available.\n\n\nDNAzymes\n\nAs mentioned above, protocols for in vitro selection of nucleic acid catalysts from random libraries have paved the way for the development of artificial RNAzymes and DNAzymes. One of the most proficient DNAzymes, the so-called 10–23 motif, was selected back in 199737 and was fully characterised in 199838 and since then has been used as a scaffold in a large number of re-selections and rational designs. In addition, novel DNAzymes were selected from fully randomised libraries. The chemical repertoire of DNAzymes is surprisingly broad, ranging from cleavage of phosphodiester, ester, and amide bonds over supporting C-C bond-forming reactions up to the repair of thymine dimers, peptide modifications, and others (excellently reviewed in 19). The recently achieved DNA-catalysed amide hydrolysis39 is a good example of the challenges in DNAzyme development. Previous selection experiments had led to DNA-catalysed DNA phosphodiester cleavage instead of the desired amide hydrolysis40, and, under conditions that deliberately avoided phosphodiester hydrolysis, no DNAzyme with activity for hydrolysis of an aliphatic amide bond was found. Instead, selection resulted in DNA catalysts that supported hydrolysis of carbonic acid esters or of aromatic amide bonds41. Only the inclusion, in the selection experiment, of nucleotide derivatives with attached protein-like functional groups allowed the identification of DNAzymes capable of aliphatic amide hydrolysis (Figure 2)39.\n\nFor more detail, see the ‘DNAzymes’ section of the main text.\n\nThere has also been some effort in elucidating the structure of DNA catalysts. A recent breakthrough is the crystal structure of an RNA-ligating deoxyribozyme at 2.8 Å resolution42. The structure gives new insight into the principles underlying DNA catalysis and allows conclusions to be drawn on the similarities and differences between RNAzymes and DNAzymes. Notably, the structure revealed that DNA can explore a wide range of conformations owing to a less restrictive sugar puckering as compared with RNA, and this feature compensates for the lack of the 2′-OH group that is present and structurally important in RNA (Figure 3).\n\nANA, arabino nucleic acid; CeNA, cyclohexene nucleic acid; FANA, 2′-fluoroarabino nucleic acid; HNA, hexitol nucleic acid.\n\nIn general, DNAzymes continue to be developed as functional modules in biosensors and computing circuits23,43–47 as well as for therapeutic use48. For example, recent progress was made in the development of variants of the 10–23 DNAzyme against hepatitis C virus49 and for the treatment of basal cell carcinoma50 as well as in DNAzyme-mediated modification of allergen-induced asthmatic responses51.\n\n\nXNAzymes\n\nVery recently, an exciting new class of nucleic acid catalysts has emerged. Artificial endonuclease and ligase enzymes composed of synthetic genetic polymers, xeno nucleic acids (XNAzymes), were selected from random libraries in a method termed ‘cross-chemistry selective enrichment by exponential amplification’ (X-SELEX)52,53. As an essential prerequisite of the experiments, a modified DNA polymerase was engineered to tolerate the XNA building blocks (triphosphates) for polymerisation54. Four different XNAs (Figure 3) were used in the selection: arabino nucleic acids (ANAs), 2′-fluoroarabino nucleic acids (FANAs), hexitol nucleic acids (HNAs), and cyclohexene nucleic acids (CeNAs), and for all of them catalytically active species were found after 10 to 20 rounds of selection52. Moreover, a FANA metalloenzyme with activity for ligation of FANA was identified, thus establishing catalysis in an entirely synthetic system53. These results have strong implications for the emergence of life on earth, underscoring the possibility that genetic polymers with backbones other than ribose may have pre-dated the emergence of RNA and the RNA world.\n\n\nRibozymes in RNA world scenarios\n\nThe discovery of ribozymes has led to a renaissance of the RNA world theory, and ever since much effort has been put into the identification of ribozymes with useful activities in a time period when life was based on RNA functioning as both genome and genome-encoded catalyst9. Thus, a number of in vitro selections aimed at the identification of RNA catalysts supporting reactions that might have been used by RNA world organisms were carried out. The synthesis of RNA certainly would have been a core activity, and ribozymes for reaction steps involved in RNA synthesis have been generated55. Recent success has been made in ribozyme-mediated triphosphorylation of RNA-5′-hydroxyl groups using cyclic trimethaphosphate as the energy source56,57, in ribozyme-mediated self-replication10,58, and in polymerisation of activated nucleotides11,12,59. In addition, other recently demonstrated activities, such as ribozyme-mediated RNA processing60,61, recombination62,63, nucleotide addition64, and self-alkylation65 (some of them illustrated in Figure 4), speak to the capacity of RNA to support a wide variety of reactions with relevance in RNA world scenarios.\n\na) self-modification, e.g. alkylation; b) 5'-terminal modification by ribozyme-supported addition of an activated building block; c) internal modification by ribozyme-supported fragment exchange; d) ribozyme-supported 5' –triphosphorylation with trimetaphosphate; e) ribozyme-supported RNA polymerization with nucleoside-2',3'-cyclic phosphates (in 3'→5'-direction) or nucleoside-5'-triphosphates (in 5'→3'-direction) as activated building blocks.\n\n\nNew catalytic motifs\n\nUntil recently, 10 classes of ribozymes existing among contemporary organisms were known, the hammerhead and hairpin ribozyme probably being the most prominent examples. The years after the discovery of these ribozymes were filled with investigations into their structures and catalytic mechanisms, and it took a rather long time until the question for additional naturally occurring ribozymes was addressed. The first of the recently discovered self-cleaving RNAs constituting the eleventh class of ribozymes is a small catalytic RNA motif, present in many species of bacteria and eukaryotes15. In keeping with the tradition of giving ribozymes names related to their secondary structure, the new motif was called twister because of its small yet complex consensus structure composed of three stems conjoined by internal and terminal loops and a two-pseudoknot tertiary fold (Figure 5)66–68. With an in vivo cleavage rate of 1000 per minute, the twister ribozyme is one of the fastest self-cleaving ribozymes, and based on biochemical experiments in conjunction with molecular dynamics simulation, a mechanism involving general acid-base catalysis by a conserved active site adenine residue has been proposed69. This is in general agreement with the mechanisms of other self-cleaving ribozymes like the hairpin or the hepatitis delta virus ribozyme, which also require an adenine residue in the active site5. However, apparently there is a striking difference: whereas in the hairpin and hepatitis delta virus ribozyme, N1 of adenine is involved in catalysis, N3 of adenine was suggested as a strong candidate to act as general base in twister ribozyme-mediated self-cleavage69. This is particularly interesting because, if indeed N3 takes this role, it would expand the mechanistic repertoire of the small endonucleolytic ribozymes.\n\nThe arrows denote the cleavage sites.\n\nHigh-throughput bioinformatics assisted the identification of additional self-cleaving candidates named twister sister, pistol, and hatchet ribozyme70, which upon in vitro characterisation were shown to indeed be ribozymes70–72. All of these new ribozymes support a transesterification reaction yielding a 5′-hydroxyl group and a 2′,3′-cyclic phosphate at the cleavage site. A recent review of the chemistry and biology of self-cleaving ribozymes referring also to the four new ribozyme classes can be found in 5.\n\n\nFuture prospects\n\nOver the years, ribozyme research and nucleic acid catalysis have remained a very exciting field with unchanged potential for new discoveries. A strong focus of current research is the uncovering and understanding of the role that ribozymes play in biological systems. The first results on the influence of self-cleaving RNA structures on genetic control are just emerging. For example, the hammerhead, the hepatitis delta virus-like, and the twister ribozyme are widespread in nature and appear in rather diverse genetic contexts14,15,73. Ribozymes have been identified in intronic regions and mobile genetic elements, suggesting a role in pre-RNA and transcript processing13,74,75. Understanding this additional level of genetic control and regulation is one of the major challenges of current and future research in this area. The ongoing development of high-throughput bioinformatic approaches will further facilitate the identification of conserved structures and the evaluation of their genetic distribution. In addition to novel genetic locations of known ribozymes, new catalytic RNA motifs may be expected to be discovered, as shown recently for the twister, twister sister, pistol, and hatchet ribozymes15,70. In the area of ribozyme engineering by rational design and in vitro/in vivo evolution, exciting results regarding new approaches for the artificial control of gene expression by allosteric ribozymes placed in non-translated regions of transcripts may be anticipated. Also, the search for ribozymes with RNA world relevant activities can be expected to continue with unbroken excitement. In this regard, the catalytic repertoire of XNAzymes52,53 will certainly be further explored.\n\nThirty-five years after the discovery of the first catalytic RNA, ribozyme research has not lost the intriguing and highly motivating flair of the first days. There are still many questions to be addressed and much is waiting to be discovered.\n\n\nAbbreviations\n\nFANA, 2′-fluoroarabino nucleic acid; SELEX, Systematic evolution of ligands by exponential enrichment; UTR, untranslated region; XNA, xeno nucleic acid.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nTuerk C, Gold L: Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science. 1990; 249(4968): 505–10. PubMed Abstract | Publisher Full Text\n\nEllington AD, Szostak JW: In vitro selection of RNA molecules that bind specific ligands. Nature. 1990; 346(6287): 818–22. PubMed Abstract | Publisher Full Text\n\nSilverman SK: Deoxyribozymes: selection design and serendipity in the development of DNA catalysts. Acc Chem Res. 2009; 42(10): 1521–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJaschke A: Artificial ribozymes and deoxyribozymes. Curr Opin Struct Biol. 2001; 11(3): 321–6. PubMed Abstract | Publisher Full Text\n\nJimenez RM, Polanco JA, Lupták A: Chemistry and Biology of Self-Cleaving Ribozymes. Trends Biochem Sci. 2015; 40(11): 648–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilson TJ, Liu Y, Lilley DMJ: Ribozymes and the mechanisms that underlie RNA catalysis. Front Chem Sci Eng. 2016; 10(2): 178–185. Publisher Full Text\n\nLiu J, Cao Z, Lu Y: Functional nucleic acid sensors. Chem Rev. 2009; 109(5): 1948–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrommer J, Appel B, Müller S: Ribozymes that can be regulated by external stimuli. Curr Opin Biotechnol. 2015; 31: 35–41. PubMed Abstract | Publisher Full Text\n\nPressman A, Blanco C, Chen IA: The RNA World as a Model System to Study the Origin of Life. Curr Biol. 2015; 25(19): R953–63. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRobertson MP, Joyce GF: Highly efficient self-replicating RNA enzymes. Chem Biol. 2014; 21(2): 238–45. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAttwater J, Wochner A, Holliger P: In-ice evolution of RNA polymerase ribozyme activity. Nat Chem. 2013; 5(12): 1011–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSczepanski JT, Joyce GF: A cross-chiral RNA polymerase ribozyme. Nature. 2014; 515(7527): 440–2. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHammann C, Luptak A, Perreault J, et al.: The ubiquitous hammerhead ribozyme. RNA. 2012; 18(5): 871–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWebb CH, Riccitelli NJ, Ruminski DJ, et al.: Widespread occurrence of self-cleaving ribozymes. Science. 2009; 326(5955): 953. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoth A, Weinberg Z, Chen AG, et al.: A widespread self-cleaving ribozyme class is revealed by bioinformatics. Nat Chem Biol. 2014; 10(1): 56–60. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGupta A, Swati D: Hammerhead Ribozymes in Archaeal Genomes: A Computational Hunt. Interdiscip Sci. 2016; 1–13. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRamesh A, Winkler WC: Metabolite-binding ribozymes. Biochim Biophys Acta. 2014; 1839(10): 989–94. PubMed Abstract | Publisher Full Text\n\nWard WL, Plakos K, DeRose VJ: Nucleic acid catalysis: metals, nucleobases, and other cofactors. Chem Rev. 2014; 114(8): 4318–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHollenstein M: DNA Catalysis: The Chemical Repertoire of DNAzymes. Molecules. 2015; 20(11): 20777–804. PubMed Abstract | Publisher Full Text\n\nMellin JR, Cossart P: Unexpected versatility in bacterial riboswitches. Trends Genet. 2015; 31(3): 150–6. PubMed Abstract | Publisher Full Text\n\nSoukup GA, Breaker RR: Engineering precision RNA molecular switches. Proc Natl Acad Sci U S A. 1999; 96(7): 3584–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFamulok M, Hartig JS, Mayer G: Functional aptamers and aptazymes in biotechnology, diagnostics, and therapy. Chem Rev. 2007; 107(9): 3715–43. PubMed Abstract | Publisher Full Text\n\nSett A, Das S, Bora U: Functional nucleic-acid-based sensors for environmental monitoring. Appl Biochem Biotechnol. 2014; 174(3): 1073–91. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTownshend B, Kennedy AB, Xiang JS, et al.: High-throughput cellular RNA device engineering. Nat Methods. 2015; 12(10): 989–94. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRehm C, Klauser B, Hartig JS: Engineering aptazyme switches for conditional gene expression in mammalian cells utilizing an in vivo screening approach. Methods Mol Biol. 2015; 1316: 127–40. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKlauser B, Atanasov J, Siewert LK, et al.: Ribozyme-based aminoglycoside switches of gene expression engineered by genetic selection in S. cerevisiae. ACS Synth Biol. 2015; 4(5): 516–25. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nNomura Y, Zhou L, Miu A, et al.: Controlling mammalian gene expression by allosteric hepatitis delta virus ribozymes. ACS Synth Biol. 2013; 2(12): 684–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFelletti M, Klauser B, Hartig JS: Screening of Genetic Switches Based on the Twister Ribozyme Motif. Methods Mol Biol. 2016; 1380: 225–39. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nWei KY, Chen YY, Smolke CD: A yeast-based rapid prototype platform for gene control elements in mammalian cells. Biotechnol Bioeng. 2013; 110(4): 1201–10. PubMed Abstract | Publisher Full Text\n\nKennedy AB, Vowles JV, d'Espaux L, et al.: Protein-responsive ribozyme switches in eukaryotic cells. Nucleic Acids Res. 2014; 42(19): 12306–21. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAusländer S, Stücheli P, Rehm C, et al.: A general design strategy for protein-responsive riboswitches in mammalian cells. Nat Methods. 2014; 11(11): 1154–60. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKlauser B, Hartig JS: An engineered small RNA-mediated genetic switch based on a ribozyme expression platform. Nucleic Acids Res. 2013; 41(10): 5542–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaragliadis A, Krajewski SS, Rehm C, et al.: Thermozymes: Synthetic RNA thermometers based on ribozyme activity. RNA Biol. 2013; 10(6): 1010–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKetzer P, Kaufmann JK, Engelhardt S, et al.: Artificial riboswitches for gene expression and replication control of DNA and RNA viruses. Proc Natl Acad Sci U S A. 2014; 111(5): E554–62. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMustoe AM, Brooks CL, Al-Hashimi HM: Hierarchy of RNA functional dynamics. Annu Rev Biochem. 2014; 83: 441–66. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRehm C, Hartig JS: In vivo screening for aptazyme-based bacterial riboswitches. Methods Mol Biol. 2014; 1111: 237–49. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSantoro SW, Joyce GF: A general purpose RNA-cleaving DNA enzyme. Proc Natl Acad Sci U S A. 1997; 94(9): 4262–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSantoro SW, Joyce GF: Mechanism and utility of an RNA-cleaving DNA enzyme. Biochemistry. 1998; 37(38): 13330–42. PubMed Abstract | Publisher Full Text\n\nZhou C, Avins JL, Klauser PC, et al.: DNA-Catalyzed Amide Hydrolysis. J Am Chem Soc. 2016; 138(7): 2106–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nChandra M, Sachdeva A, Silverman SK: DNA-catalyzed sequence-specific hydrolysis of DNA. Nat Chem Biol. 2009; 5(10): 718–20. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBrandsen BM, Hesser AR, Castner MA, et al.: DNA-catalyzed hydrolysis of esters and aromatic amides. J Am Chem Soc. 2013; 135(43): 16014–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPonce-Salvatierra A, Wawrzyniak-Turek K, Steuerwald U, et al.: Crystal structure of a DNA catalyst. Nature. 2016; 529(7585): 231–4. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nXiang Y, Lu Y: DNA as sensors and imaging agents for metal ions. Inorg Chem. 2014; 53(4): 1925–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Y, Tang L, Zeng G, et al.: Current progress in biosensors for heavy metal ions based on DNAzymes/DNA molecules functionalized nanostructures: A review. Sens Actuators B Chem. 2016; 223: 280–94. Publisher Full Text\n\nGong L, Zhao Z, Lv YF, et al.: DNAzyme-based biosensors and nanodevices. Chem Commun (Camb). 2015; 51(6): 979–95. PubMed Abstract | Publisher Full Text\n\nHuang PJ, Liu M, Liu J: Functional nucleic acids for detecting bacteria. Rev Anal Chem. 2013; 32(1): 77–89. Publisher Full Text\n\nOrbach R, Willner B, Willner I: Catalytic nucleic acids (DNAzymes) as functional units for logic gates and computing circuits: from basic principles to practical applications. Chem Commun (Camb). 2015; 51(20): 4144–60. PubMed Abstract | Publisher Full Text\n\nKarnati HK, Yalagala RS, Undi R, et al.: Therapeutic potential of siRNA and DNAzymes in cancer. Tumour Biol. 2014; 35(10): 9505–21. PubMed Abstract | Publisher Full Text\n\nRobaldo L, Berzal-Herranz A, Montserrat JM, et al.: Activity of core-modified 10–23 DNAzymes against HCV. ChemMedChem. 2014; 9(9): 2172–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nCho EA, Moloney FJ, Cai H, et al.: Safety and tolerability of an intratumorally injected DNAzyme, Dz13, in patients with nodular basal-cell carcinoma: a phase 1 first-in-human trial (DISCOVER). Lancet. 2013; 381(9880): 1835–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrug N, Hohlfeld JM, Kirsten AM, et al.: Allergen-induced asthmatic responses modified by a GATA3-specific DNAzyme. N Engl J Med. 2015; 372(21): 1987–95. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTaylor AI, Holliger P: Directed evolution of artificial enzymes (XNAzymes) from diverse repertoires of synthetic genetic polymers. Nat Protoc. 2015; 10(10): 1625–42. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTaylor AI, Pinheiro VB, Smola MJ, et al.: Catalysts from synthetic genetic polymers. Nature. 2015; 518(7539): 427–30. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPinheiro VB, Taylor AI, Cozens C, et al.: Synthetic genetic polymers capable of heredity and evolution. Science. 2012; 336(6079): 341–4. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMartin LL, Unrau PJ, Müller UF: RNA synthesis by in vitro selected ribozymes for recreating an RNA world. Life (Basel). 2015; 5(1): 247–68. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMoretti JE, Müller UF: A ribozyme that triphosphorylates RNA 5'-hydroxyl groups. Nucleic Acids Res. 2014; 42(7): 4767–78. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nDolan GF, Akoopie A, Müller UF: A Faster Triphosphorylation Ribozyme. PLoS One. 2015; 10(11): e0142559. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFerretti AC, Joyce GF: Kinetic properties of an RNA enzyme that undergoes self-sustained exponential amplification. Biochemistry. 2013; 52(7): 1227–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWochner A, Attwater J, Coulson A, et al.: Ribozyme-catalyzed transcription of an active ribozyme. Science. 2011; 332(6026): 209–12. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPetkovic S, Badelt S, Block S, et al.: Sequence-controlled RNA self-processing: computational design, biochemical analysis, and visualization by AFM. RNA. 2015; 21(7): 1249–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmini ZN, Olson KE, Müller UF: Spliceozymes: ribozymes that remove introns from pre-mRNAs in trans. PLoS One. 2014; 9(7): e101932. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHieronymus R, Godehard SP, Balke D, et al.: Hairpin ribozyme mediated RNA recombination. Chem Commun (Camb). 2016; 52(23): 4365–8. PubMed Abstract | Publisher Full Text\n\nLehman N, Vaidya N, Yeates JA: RNA-directed recombination of RNA in vitro. Methods Mol Biol. 2015; 1240: 27–37. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMutschler H, Holliger P: Non-canonical 3'-5' extension of RNA with prebiotically plausible ribonucleoside 2',3'-cyclic phosphates. J Am Chem Soc. 2014; 136(14): 5193–6. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSharma AK, Plant JJ, Rangel AE, et al.: Fluorescent RNA labeling using self-alkylating ribozymes. ACS Chem Biol. 2014; 9(8): 1680–4. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nLiu Y, Wilson TJ, McPhee SA, et al.: Crystal structure and mechanistic investigation of the twister ribozyme. Nat Chem Biol. 2014; 10(9): 739–44. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRen A, Kosutic M, Rajashankar KR, et al.: In-line alignment and Mg2+ coordination at the cleavage site of the env22 twister ribozyme. Nat Commun. 2014; 5: 5534. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nEiler D, Wang J, Steitz TA: Structural basis for the fast self-cleavage reaction catalyzed by the twister ribozyme. Proc Natl Acad Sci U S A. 2014; 111(36): 13028–33. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGaines CS, York DM: Ribozyme Catalysis with a Twist: Active State of the Twister Ribozyme in Solution Predicted from Molecular Simulation. J Am Chem Soc. 2016; 138(9): 3058–65. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWeinberg Z, Kim PB, Chen TH, et al.: New classes of self-cleaving ribozymes revealed by comparative genomics analysis. Nat Chem Biol. 2015; 11(8): 606–10. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHarris KA, Lünse CE, Li S, et al.: Biochemical analysis of pistol self-cleaving ribozymes. RNA. 2015; 21(11): 1852–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLi S, Lunse CE, Harris KA, et al.: Biochemical analysis of hatchet self-cleaving ribozymes. RNA. 2015; 21(11): 1845–51. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPerreault J, Weinberg Z, Roth A, et al.: Identification of hammerhead ribozymes in all domains of life reveals novel structural variations. PLoS Comput Biol. 2011; 7(5): e1002031. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde la Peña M, García-Robles I: Intronic hammerhead ribozymes are ultraconserved in the human genome. EMBO Rep. 2010; 11(9): 711–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCervera A, De La Peña M: Eukaryotic penelope-like retroelements encode hammerhead ribozyme motifs. Mol Biol Evol. 2014; 31(11): 2941–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation"
}
|
[
{
"id": "14599",
"date": "27 Jun 2016",
"name": "Jörg Hartig",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14603",
"date": "27 Jun 2016",
"name": "Marcel Hollenstein",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1511
|
https://f1000research.com/articles/5-1510/v1
|
27 Jun 16
|
{
"type": "Review",
"title": "Extracoporeal photopheresis treatment of acute graft-versus-host disease following allogeneic haematopoietic stem cell transplantation",
"authors": [
"Aisling M. Flinn",
"Andrew R. Gennery",
"Andrew R. Gennery"
],
"abstract": "Acute graft-versus-host disease (aGvHD) continues to be a major obstacle to allogeneic haematopoietic stem cell transplantation. Thymic damage secondary to aGvHD along with corticosteroids and other non-selective T lymphocyte-suppressive agents used in the treatment of aGvHD concurrently impair thymopoiesis and negatively impact on immunoreconstitution of the adaptive immune compartment and ultimately adversely affect clinical outcome. Extracorporeal photopheresis (ECP) is an alternative therapeutic strategy that appears to act in an immunomodulatory fashion, potentially involving regulatory T lymphocytes and dendritic cells. By promoting immune tolerance and simultaneously avoiding systemic immunosuppression, ECP could reduce aGvHD and enable a reduction in other immunosuppression, allowing thymic recovery, restoration of normal T lymphopoiesis, and complete immunoreconstitution with improved clinical outcome. Although the safety and efficacy of ECP has been demonstrated, further randomised controlled studies are needed as well as elucidation of the underlying mechanisms responsible and the effect of ECP on thymic recovery.",
"keywords": [
"Acute graft-versus-host disease",
"Acute graft-versus-host disease",
"stem cell transplantation",
"aGvHD",
"Extracorporeal photopheresis"
],
"content": "Introduction\n\nAllogeneic haematopoietic stem cell transplantation (HSCT) is used to treat malignant and non-malignant haematological conditions1. In primary immunodeficiency, the aim following HSCT is to achieve complete and long-lasting immunoreconstitution (IR) with a diverse T cell receptor (TCR) repertoire, providing adequate adaptive T lymphocyte immunity2. Delayed or persisting immunodeficiency is associated with significant morbidity and mortality with increased risk of infection, relapse, and development of secondary malignancies3,4. Potential strategies to boost thymic function and promote faster and complete IR, particularly in older patients who exhibit reduced thymic function inherently due to aging, have garnered much interest to improve patient outcome. Such approaches include the use of Fgf7 or sex steroid hormone inhibition, which have been shown to protect thymic epithelial cells (TECs) and improve thymopoiesis in experimental models5.\n\n\nEffect of graft-versus-host disease on T lymphocyte immunoreconstitution\n\nConditioning given prior to HSCT results in an inevitable period of aplasia with obliteration of innate and adaptive immune responses, subjecting the patient to a period of increased risk of infection and other complications until the stem cells engraft and reconstitution of the immune system compartments ensues. Rebuilding of innate immunity, including monocytes, granulocytes, and epithelial barriers, occurs relatively quickly following HSCT, providing protection against bacterial and fungal infections6. In contrast, T lymphocyte reconstitution is lengthier and more complex, involving two pathways6–8. Peripheral thymic-independent expansion of surviving host T lymphocytes and/or transferred donor T lymphocytes provides a degree of immediate T lymphocyte immunity but of limited diversity and permanency5. Complete IR following lymphodepletion requires durable de novo thymic regeneration of naïve T lymphocytes from donor progenitor cells with a broad TCR repertoire, which requires a functioning and structurally intact thymus8,9. These naïve T lymphocytes (termed recent thymic emigrants [RTEs]) can be measured quantitatively by identification of surface markers such as CD45RA and CD31 using flow cytometry and by determination of TCR excision circle (TREC) levels. TRECs are circular pieces of DNA produced as a consequence of TCR α and β chain formation, and quantification of TREC content in T lymphocytes provides a practical and accepted measurement of thymic output10. The quality of the T lymphocyte compartment can be assessed by measuring TCR diversity, as this is almost completely reflective of the naïve T lymphocyte population11. This can be done using flow cytometry, spectratyping of the complementarity determining region 3 (CDR3), and nucleotide sequencing. Flow cytometry is widely available and cheaper and results can be obtained quickly12. Spectratyping analyses the lengths of the hypervariable region CDR3 in each Vβ family using real-time polymerase chain reaction13,14. Compared to flow cytometry, spectratyping provides more detailed resolution of TCR diversity; however, there is no accepted single standardised method of analysing data at present, and this technique gives equal weighting to all Vβ families measured, independent of how many genes they contain13. Nucleotide sequencing of DNA CDR3 regions provides even more in-depth analysis but is expensive and, although evolving, is not widely available at present15. Thymic damage disrupts normal T lymphocyte ontogeny, resulting in reduced export of RTEs and a distorted TCR repertoire, negatively impacting on IR and clinical outcome5,16–18.\n\nGraft-versus-host disease (GvHD) is a leading cause of post-HSCT mortality19,20. Acute (a)GvHD is mediated by alloreactive mature donor T lymphocytes, which attack disparate recipient antigens, resulting in a harmful inflammatory response and tissue injury21. Elucidation of aGvHD pathophysiology is based on experimental models20: (1) damage to host tissue by conditioning regimens, underlying disease, and/or infections increases pro-inflammatory cytokines activating host antigen-presenting cells (APCs); (2) donor T lymphocytes recognise the disparate alloantigens on activated host APCs and become activated, proliferate, differentiate, produce further inflammatory cytokines, and migrate to target organs; (3) effector cells, primarily cytotoxic T lymphocytes and natural killer (NK) cells, and soluble effectors cause apoptosis of target cells.\n\nAlthough aGvHD principally involves the skin, gastrointestinal tract, and liver, the thymus is also a primary target, resulting in disruption of thymic architecture with loss of cortico-medullary demarcation, alteration of TEC subpopulations, and depletion of thymocytes22–24. The precise mechanisms behind aGvHD-induced thymic injury in humans remain incompletely understood, but experimental models have helped delineate the underlying cellular and molecular mechanisms22. TECs are initiators and targets of thymic aGvHD, capable of activating alloreactive donor T lymphocytes independently of APCs, leading to secretion of interferon gamma (IFNγ) and triggering signal transducer and activator of transcription 1 (STAT1)-induced apoptosis of cortical and medullary TECs9. The resulting disruption of architecture and organisation of the thymic microenvironment with thymic atrophy disturbs the normal signalling required for immature thymocyte development, particularly at the triple-negative proliferative stage and with increased apoptosis of double-positive cells22,25,26, resulting in impaired lymphopoiesis and reduced thymic export (Figure 1)11,27. Acute GvHD also impairs the thymic-independent pathway with reduced expansion of transferred mature donor T lymphocytes, possibly due to loss of peripheral T lymphocyte niches28.\n\nThymic damage occurs secondary to allogeneic T lymphocytotoxicity during aGvHD, corticosteroid-mediated damage, and other non-selective T lymphocyte-suppressive agents used in the treatment of aGvHD, causing impaired thymopoiesis (A), with reduced thymic export and a distorted T cell receptor (TCR) repertoire with potentially autoreactive thymocytes escaping negative selection (B). ECP, by promoting immune tolerance and enabling reduction and cessation of conventional immunosuppression, may allow thymic recovery, resumption of normal thymopoiesis, and complete and long-lasting immunoreconstitution post-haematopoietic stem cell transplantation (C). Abbreviations: Treg, regulatory T lymphocyte; DP, double positive.\n\nA distorted TCR repertoire is observed in patients with aGvHD10. Disparate donor and recipient major histocompatibility complex (MHC) complexes disturb thymic positive and negative selection, impacting on TCR selection, resulting in thymocytes escaping negative selection, and increasing the survival of autoreactive T lymphocytes29–32. Thus, aGvHD is detrimental to the quantity and quality of T lymphocyte recovery. The thymus is particularly sensitive to aGvHD, with thymic output being significantly affected, even in grade 1 disease11. Subclinical thymic aGvHD may have an underappreciated adverse effect on the reconstitution of adaptive immunity, causing ongoing infections and incomplete IR post-HSCT.\n\n\nCorticosteroid treatment of acute graft-versus-host disease\n\nCorticosteroids, with potent immunosuppressive and anti-inflammatory effects, are the first-line treatment for aGvHD, but a complete response is witnessed in only 25–50% of patients33. Short, intensive courses of corticosteroids induce thymic involution, causing a profound reduction in naïve T lymphocyte production, although with complete recovery following cessation34. However, the precise effects in human thymus and of long-term corticosteroid use are unknown. There is no consensus for second-line therapy for steroid-dependent/-refractory aGvHD, which usually involves the intensification of systemic immunosuppression with a plethora of therapeutic agents that non-selectively target T lymphocytes35,36. Second-line options include mycophenolate mofetil, anti-tumour necrosis factor alpha antibodies, or mammalian target of rapamycin (mTOR) inhibitors. The use of mesenchymal stromal cells has also been advocated, with mixed success, in part because the product is a cellular therapy and it is difficult to ensure consistency of the cellular content37–39. Acute GvHD and immunosuppressive treatment concurrently impair thymopoiesis, subjecting patients to further risk of infection, relapse, and development of secondary malignancies, as well as associated toxicity40,41. A targeted therapy for aGvHD without systemic immunosuppression and that allows thymic recovery is needed42.\n\n\nExtracorporeal photopheresis\n\nExtracorporeal photopheresis (ECP) exposes apheresed mononuclear cells to 8-methoxypsoralen and UVA radiation, with re-infusion of photoactivated cells into the patient43. This induces DNA damage and apoptosis of exposed cells, with activated T lymphocytes preferentially affected44,45. As only 5-10% of lymphocytes are exposed during the procedure, which is insufficient to explain the effects of ECP, it is speculated that the apoptotic cells have indirect immunomodulatory actions on other immunocompetent cells43. These immunomodulatory mechanisms are poorly understood, but generation of regulatory T lymphocytes (Tregs), alteration of cytokine patterns, and modulation of dendritic cells (DCs) appear to be fundamental46–52.\n\nThe modulation of DCs includes increased number due to differentiation of ECP-exposed monocytes53,54 and stimulation of a DC-tolerogenic state upon phagocytosis of apoptosed cells, characterised by down-regulation of maturation markers and co-stimulatory molecules and increased secretion of anti-inflammatory cytokines, particularly interleukin-1055–60. Upon interaction with T lymphocytes, tolerogenic DCs can induce T lymphocyte anergy or apoptosis or stimulate Treg production58,61. In aGvHD, DCs, as the major APC, present disparate host antigens to donor T lymphocytes, propagating the pathway of cellular injury. Inducing a DC-tolerogenic state and dampening T lymphocyte activation could attenuate the trigger for aGvHD. The modulation of DC number and function may be a central mechanism of ECP. Tregs are essential in maintaining self-tolerance, down-regulating immune responses, and limiting inflammation that may be harmful to the host and contribute to the mechanism of ECP62–67.\n\nThe unique advantage of ECP as a therapy is lack of global immunosuppression but preservation of the graft-versus-leukaemia effect68. Promoting immune tolerance, with selective down-regulation of immune stimulation, could reduce aGvHD and enable a reduction in other immunosuppression, facilitating thymic recovery, restoration of normal T lymphopoiesis, and complete IR (Figure 1) with improved clinical outcome as ability to fight infections improves and risk of secondary malignancy or relapse diminishes. It is well tolerated with few adverse effects, and reports of clinical efficacy are impressive69–77. Whilst the immune-sparing effects of ECP have been demonstrated78,79, further randomised controlled studies are needed as well as investigation of the effect of ECP on thymic recovery. Further elucidation of the underlying mechanisms at play, as well as the optimal treatment schedule, is required to ascertain fully the role of ECP in aGvHD treatment.\n\n\nAbbreviations\n\naGvHD, acute graft-versus-host disease; APC, antigen-presenting cell; CDR3, complementarity determining region 3; DC, dendritic cell; ECP, extracorporeal photopheresis; HSCT, allogeneic hematopoietic stem cell transplant; IFNγ, interferon gamma; IR, immunoreconstitution; MHC, major histocompatibility complex; mTOR, mammalian target of rapamycin; NK, natural killer; STAT1, signal transducer and activator of transcription 1; TCR, T cell receptor; TEC, thymic epithelial cell; TREC, T cell receptor excision circle; Treg, regulatory T lymphocyte.",
"appendix": "Competing interests\n\n\n\nThe author(s) declared that they have no competing interests.\n\n\nGrant information\n\nAisling Flinn is funded by the Bubble Foundation UK.\n\n\nReferences\n\nLjungman P, Bregni M, Brune M, et al.: Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: current practice in Europe 2009. Bone Marrow Transplant. 2010; 45(2): 219–34. PubMed Abstract | Publisher Full Text\n\nBurroughs L, Woolfrey A: Hematopoietic cell transplantation for treatment of primary immune deficiencies. Cell Ther Transplant. 2010; 2(8): PubMed Abstract | Publisher Full Text | Free Full Text\n\nBosch M, Khan FM, Storek J: Immune reconstitution after hematopoietic cell transplantation. Curr Opin Hematol. 2012; 19(4): 324–35. PubMed Abstract | Publisher Full Text\n\nAntin JH: Immune reconstitution: the major barrier to successful stem cell transplantation. Biol Blood Marrow Transplant. 2005; 11(2 Suppl 2): 43–5. PubMed Abstract | Publisher Full Text\n\nKrenger W, Holländer GA: The role of the thymus in allogeneic hematopoietic stem cell transplantation. Swiss Med Wkly. 2010; 140: w13051. PubMed Abstract | Publisher Full Text\n\nStorek J, Geddes M, Khan F, et al.: Reconstitution of the immune system after hematopoietic stem cell transplantation in humans. Semin Immunopathol. 2008; 30(4): 425–37. PubMed Abstract | Publisher Full Text\n\nKrenger W, Blazar BR, Holländer GA: Thymic T-cell development in allogeneic stem cell transplantation. Blood. 2011; 117(25): 6768–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDumont-Girard F, Roux E, van Lier RA, et al.: Reconstitution of the T-cell compartment after bone marrow transplantation: restoration of the repertoire by thymic emigrants. Blood. 1998; 92(11): 4464–71. PubMed Abstract\n\nHauri-Hohl MM, Keller MP, Gill J, et al.: Donor T-cell alloreactivity against host thymic epithelium limits T-cell development after bone marrow transplantation. Blood. 2007; 109(9): 4080–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClave E, Busson M, Douay C, et al.: Acute graft-versus-host disease transiently impairs thymic output in young patients after allogeneic hematopoietic stem cell transplantation. Blood. 2009; 113(25): 6477–84. PubMed Abstract | Publisher Full Text\n\nToubert A, Glauzy S, Douay C, et al.: Thymus and immune reconstitution after allogeneic hematopoietic stem cell transplantation in humans: never say never again. Tissue Antigens. 2012; 79(2): 83–9. PubMed Abstract | Publisher Full Text\n\nLangerak AW, van Den Beemd R, Wolvers-Tettero IL, et al.: Molecular and flow cytometric analysis of the Vbeta repertoire for clonality assessment in mature TCRalphabeta T-cell proliferations. Blood. 2001; 98(1): 165–73. PubMed Abstract | Publisher Full Text\n\nMemon SA, Sportés C, Flomerfelt FA, et al.: Quantitative analysis of T cell receptor diversity in clinical samples of human peripheral blood. J Immunol Methods. 2012; 375(1–2): 84–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGenevée C, Diu A, Nierat J, et al.: An experimentally validated panel of subfamily-specific oligonucleotide primers (V alpha 1-w29/V beta 1-w24) for the study of human T cell receptor variable V gene segment usage by polymerase chain reaction. Eur J Immunol. 1992; 22(5): 1261–9. PubMed Abstract | Publisher Full Text\n\nSix A, Mariotti-Ferrandiz ME, Chaara W, et al.: The past, present, and future of immune repertoire biology - the rise of next-generation repertoire analysis. Front Immunol. 2013; 4: 413. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHazenberg MD, Otto SA, de Pauw ES, et al.: T-cell receptor excision circle and T-cell dynamics after allogeneic stem cell transplantation are related to clinical events. Blood. 2002; 99(9): 3449–53. PubMed Abstract | Publisher Full Text\n\nClave E, Lisini D, Douay C, et al.: A low thymic function is associated with leukemia relapse in children given T-cell-depleted HLA-haploidentical stem cell transplantation. Leukemia. 2012; 26(8): 1886–8. PubMed Abstract | Publisher Full Text\n\nOlkinuora H, Talvensaari K, Kaartinen T, et al.: T cell regeneration in pediatric allogeneic stem cell transplantation. Bone Marrow Transplant. 2007; 39(3): 149–56. PubMed Abstract | Publisher Full Text\n\nWelniak LA, Blazar BR, Murphy WJ: Immunobiology of allogeneic hematopoietic stem cell transplantation. Annu Rev Immunol. 2007; 25: 139–70. PubMed Abstract | Publisher Full Text\n\nFerrara JL, Levine JE, Reddy P, et al.: Graft-versus-host disease. Lancet. 2009; 373(9674): 1550–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSung AD, Chao NJ: Concise review: acute graft-versus-host disease: immunobiology, prevention, and treatment. Stem Cells Transl Med. 2013; 2(1): 25–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrenger W, Rossi S, Piali L, et al.: Thymic atrophy in murine acute graft-versus-host disease is effected by impaired cell cycle progression of host pro-T and pre-T cells. Blood. 2000; 96(1): 347–54. PubMed Abstract\n\nKrenger W, Holländer GA: The immunopathology of thymic GVHD. Semin Immunopathol. 2008; 30(4): 439–56. PubMed Abstract | Publisher Full Text\n\nSeemayer TA, Lapp WS, Bolande RP: Thymic involution in murine graft-versus-host reaction. Epithelial injury mimicking human thymic dysplasia. Am J Pathol. 1977; 88(1): 119–34. PubMed Abstract | Free Full Text\n\nKrenger W, Rossi S, Holländer GA: Apoptosis of thymocytes during acute graft-versus-host disease is independent of glucocorticoids. Transplantation. 2000; 69(10): 2190–3. PubMed Abstract\n\nKrenger W, Holländer GA: The thymus in GVHD pathophysiology. Best Pract Res Clin Haematol. 2008; 21(2): 119–28. PubMed Abstract | Publisher Full Text\n\nWeinberg K, Blazar BR, Wagner JE, et al.: Factors affecting thymic function after allogeneic hematopoietic stem cell transplantation. Blood. 2001; 97(5): 1458–66. PubMed Abstract | Publisher Full Text\n\nDulude G, Roy DC, Perreault C: The effect of graft-versus-host disease on T cell production and homeostasis. J Exp Med. 1999; 189(8): 1329–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHolländer GA, Widmer B, Burakoff SJ: Loss of normal thymic repertoire selection and persistence of autoreactive T cells in graft vs host disease. J Immunol. 1994; 152(4): 1609–17. PubMed Abstract\n\nTivol E, Komorowski R, Drobyski WR: Emergent autoimmunity in graft-versus-host disease. Blood. 2005; 105(12): 4885–91. PubMed Abstract | Publisher Full Text\n\nTeshima T, Reddy P, Liu C, et al.: Impaired thymic negative selection causes autoimmune graft-versus-host disease. Blood. 2003; 102(2): 429–35. PubMed Abstract | Publisher Full Text\n\nSakoda Y, Hashimoto D, Asakura S, et al.: Donor-derived thymic-dependent T cells cause chronic graft-versus-host disease. Blood. 2007; 109(4): 1756–64. PubMed Abstract | Publisher Full Text\n\nMacMillan ML, Weisdorf DJ, Wagner JE, et al.: Response of 443 patients to steroids as primary therapy for acute graft-versus-host disease: comparison of grading systems. Biol Blood Marrow Transplant. 2002; 8(7): 387–94. PubMed Abstract | Publisher Full Text\n\nKong FK, Chen CL, Cooper MD: Reversible disruption of thymic function by steroid treatment. J Immunol. 2002; 168(12): 6500–5. PubMed Abstract | Publisher Full Text\n\nMartin PJ, Rizzo JD, Wingard JR, et al.: First- and second-line systemic treatment of acute graft-versus-host disease: recommendations of the American Society of Blood and Marrow Transplantation. Biol Blood Marrow Transplant. 2012; 18(8): 1150–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDignan FL, Clark A, Amrolia P, et al.: Diagnosis and management of acute graft-versus-host disease. Br J Haematol. 2012; 158(1): 30–45. PubMed Abstract | Publisher Full Text\n\nRizk M, Monaghan M, Shorr R, et al.: Heterogeneity in Studies of Mesenchymal Stromal Cells to Treat or Prevent Graft-versus-Host Disease: A Scoping Review of the Evidence. Biol Blood Marrow Transplant. 2016; pii: S1083-8791(16)30029-5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKim N, Im KI, Lim JY, et al.: Mesenchymal stem cells for the treatment and prevention of graft-versus-host disease: experiments and practice. Ann Hematol. 2013; 92(10): 1295–308. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPrasad VK, Lucas KG, Kleiner GI, et al.: Efficacy and safety of ex vivo cultured adult human mesenchymal stem cells (Prochymal™) in pediatric patients with severe refractory acute graft-versus-host disease in a compassionate use study. Biol Blood Marrow Transplant. 2011; 17(4): 534–41. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSeggewiss R, Einsele H: Immune reconstitution after allogeneic transplantation and expanding options for immunomodulation: an update. Blood. 2010; 115(19): 3861–8. PubMed Abstract | Publisher Full Text\n\nFry TJ, Mackall CL: Immune reconstitution following hematopoietic progenitor cell transplantation: challenges for the future. Bone Marrow Transplant. 2005; 35(Suppl 1): S53–7. PubMed Abstract | Publisher Full Text\n\nKitko CL, Levine JE: Extracorporeal photopheresis in prevention and treatment of acute GVHD. Transfus Apher Sci. 2015; 52(2): 151–6. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKnobler R, Barr ML, Couriel DR, et al.: Extracorporeal photopheresis: past, present, and future. J Am Acad Dermatol. 2009; 61(4): 652–65. PubMed Abstract | Publisher Full Text\n\nHannani D, Merlin E, Gabert F, et al.: Photochemotherapy induces a faster apoptosis of alloreactive activated T cells than of nonalloreactive resting T cells in graft versus host disease. Transplantation. 2010; 90(11): 1232–8. PubMed Abstract | Publisher Full Text\n\nTambur AR, Ortegel JW, Morales A, et al.: Extracorporeal photopheresis induces lymphocyte but not monocyte apoptosis. Transplant Proc. 2000; 32(4): 747–8. PubMed Abstract | Publisher Full Text\n\nXia CQ, Campbell KA, Clare-Salzler MJ: Extracorporeal photopheresis-induced immune tolerance: a focus on modulation of antigen-presenting cells and induction of regulatory T cells by apoptotic cells. Curr Opin Organ Transplant. 2009; 14(4): 338–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGorgun G, Miller KB, Foss FM: Immunologic mechanisms of extracorporeal photochemotherapy in chronic graft-versus-host disease. Blood. 2002; 100(3): 941–7. PubMed Abstract | Publisher Full Text\n\nBladon J, Taylor PC: Extracorporeal photopheresis: a focus on apoptosis and cytokines. J Dermatol Sci. 2006; 43(2): 85–94. PubMed Abstract | Publisher Full Text\n\nGeorge JF, Gooden CW, Guo L, et al.: Role for CD4+CD25+ T cells in inhibition of graft rejection by extracorporeal photopheresis. J Heart Lung Transplant. 2008; 27(6): 616–22. PubMed Abstract | Publisher Full Text\n\nDi Renzo M, Rubegni P, Pasqui AL, et al.: Extracorporeal photopheresis affects interleukin (IL)-10 and IL-12 production by monocytes in patients with chronic graft-versus-host disease. Br J Dermatol. 2005; 153(1): 59–65. PubMed Abstract | Publisher Full Text\n\nDieterlen MT, Bittner HB, Pierzchalski A, et al.: Immunological monitoring of extracorporeal photopheresis after heart transplantation. Clin Exp Immunol. 2014; 176(1): 120–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHeshmati F: Updating ECP action mechanisms. Transfus Apher Sci. 2014; 50(3): 330–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBerger C, Hoffmann K, Vasquez JG, et al.: Rapid generation of maturationally synchronized human dendritic cells: contribution to the clinical efficacy of extracorporeal photochemotherapy. Blood. 2010; 116(23): 4838–47. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nEdelson RL: Mechanistic insights into extracorporeal photochemotherapy: efficient induction of monocyte-to-dendritic cell maturation. Transfus Apher Sci. 2014; 50(3): 322–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHoltick U, Marshall SR, Wang XN, et al.: Impact of psoralen/UVA-treatment on survival, activation, and immunostimulatory capacity of monocyte-derived dendritic cells. Transplantation. 2008; 85(5): 757–66. PubMed Abstract | Publisher Full Text\n\nSpisek R, Gasova Z, Bartunkova J: Maturation state of dendritic cells during the extracorporeal photopheresis and its relevance for the treatment of chronic graft-versus-host disease. Transfusion. 2006; 46(1): 55–65. PubMed Abstract | Publisher Full Text\n\nDi Renzo M, Sbano P, De Aloe G, et al.: Extracorporeal photopheresis affects co-stimulatory molecule expression and interleukin-10 production by dendritic cells in graft-versus-host disease patients. Clin Exp Immunol. 2008; 151(3): 407–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCapitini CM, Davis JP, Larabee SM, et al.: Extracorporeal photopheresis attenuates murine graft-versus-host disease via bone marrow-derived interleukin-10 and preserves responses to dendritic cell vaccination. Biol Blood Marrow Transplant. 2011; 17(6): 790–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nFutterleib JS, Feng H, Tigelaar RE, et al.: Activation of GILZ gene by photoactivated 8-methoxypsoralen: potential role of immunoregulatory dendritic cells in extracorporeal photochemotherapy. Transfus Apher Sci. 2014; 50(3): 379–87. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLegitimo A, Consolini R, Failli A, et al.: In vitro treatment of monocytes with 8-methoxypsolaren and ultraviolet A light induces dendritic cells with a tolerogenic phenotype. Clin Exp Immunol. 2007; 148(3): 564–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorelli AE, Thomson AW: Tolerogenic dendritic cells and the quest for transplant tolerance. Nat Rev Immunol. 2007; 7(8): 610–21. PubMed Abstract | Publisher Full Text\n\nGatza E, Rogers CE, Clouthier SG, et al.: Extracorporeal photopheresis reverses experimental graft-versus-host disease through regulatory T cells. Blood. 2008; 112(4): 1515–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLamioni A, Parisi F, Isacchi G, et al.: The immunological effects of extracorporeal photopheresis unraveled: induction of tolerogenic dendritic cells in vitro and regulatory T cells in vivo. Transplantation. 2005; 79(7): 846–50. PubMed Abstract | Publisher Full Text\n\nBiagi E, Di Biaso I, Leoni V, et al.: Extracorporeal photochemotherapy is accompanied by increasing levels of circulating CD4+CD25+GITR+Foxp3+CD62L+ functional regulatory T-cells in patients with graft-versus-host disease. Transplantation. 2007; 84(1): 31–9. PubMed Abstract | Publisher Full Text\n\nLorenz K, Rommel K, Mani J, et al.: Modulation of lymphocyte subpopulations by extracorporeal photopheresis in patients with acute graft-versus-host disease or graft rejection. Leuk Lymphoma. 2015; 56(3): 671–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMaeda A, Schwarz A, Kernebeck K, et al.: Intravenous infusion of syngeneic apoptotic cells by photopheresis induces antigen-specific regulatory T cells. J Immunol. 2005; 174(10): 5968–76. PubMed Abstract | Publisher Full Text\n\nSchmitt S, Johnson TS, Karakhanova S, et al.: Extracorporeal photophoresis augments function of CD4+CD25+FoxP3+ regulatory T cells by triggering adenosine production. Transplantation. 2009; 88(3): 411–6. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMarshall SR: Technology insight: ECP for the treatment of GvHD--can we offer selective immune control without generalized immunosuppression? Nat Clin Pract Oncol. 2006; 3(6): 302–14. PubMed Abstract | Publisher Full Text\n\nCalore E, Marson P, Pillon M, et al.: Treatment of Acute Graft-versus-Host Disease in Childhood with Extracorporeal Photochemotherapy/Photopheresis: The Padova Experience. Biol Blood Marrow Transplant. 2015; 21(11): 1963–72. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBruserud Ø, Tvedt TH, Paulsen PQ, et al.: Extracorporeal photopheresis (photochemotherapy) in the treatment of acute and chronic graft versus host disease: immunological mechanisms and the results from clinical studies. Cancer Immunol Immunother. 2014; 63(8): 757–77. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nHeshmati F: Extra corporeal photo chemotherapy (ECP) in acute and chronic GVHD. Transfus Apher Sci. 2010; 43(2): 211–5. PubMed Abstract | Publisher Full Text\n\nDas-Gupta E, Dignan F, Shaw B, et al.: Extracorporeal photopheresis for treatment of adults and children with acute GVHD: UK consensus statement and review of published literature. Bone Marrow Transplant. 2014; 49(10): 1251–8. PubMed Abstract | Publisher Full Text\n\nRutella S, Valentini CG, Ceccarelli S, et al.: Extracorporeal photopheresis for paediatric patients experiencing graft-versus-host disease (GVHD). Transfus Apher Sci. 2014; 50(3): 340–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nAbu-Dalle I, Reljic T, Nishihori T, et al.: Extracorporeal photopheresis in steroid-refractory acute or chronic graft-versus-host disease: results of a systematic review of prospective studies. Biol Blood Marrow Transplant. 2014; 20(11): 1677–86. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMessina C, Locatelli F, Lanino E, et al.: Extracorporeal photochemotherapy for paediatric patients with graft-versus-host disease after haematopoietic stem cell transplantation. Br J Haematol. 2003; 122(1): 118–27. PubMed Abstract | Publisher Full Text\n\nBerger M, Pessolano R, Albiani R, et al.: Extracorporeal photopheresis for steroid resistant graft versus host disease in pediatric patients: a pilot single institution report. J Pediatr Hematol Oncol. 2007; 29(10): 678–87. PubMed Abstract | Publisher Full Text\n\nHart JW, Shiue LH, Shpall EJ, et al.: Extracorporeal photopheresis in the treatment of graft-versus-host disease: evidence and opinion. Ther Adv Hematol. 2013; 4(5): 320–34. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nUssowicz M, Musial J, Mielcarek M, et al.: Steroid-sparing effect of extracorporeal photopheresis in the therapy of graft-versus-host disease after allogeneic hematopoietic stem cell transplantation. Transplant Proc. 2013; 45(9): 3375–80. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nDignan FL, Aguilar S, Scarisbrick JJ, et al.: Impact of extracorporeal photopheresis on skin scores and quality of life in patients with steroid-refractory chronic GVHD. Bone Marrow Transplant. 2014; 49(5): 704–8. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation"
}
|
[
{
"id": "14598",
"date": "27 Jun 2016",
"name": "Frank J. Staal",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14597",
"date": "27 Jun 2016",
"name": "Jan Stary",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1510
|
https://f1000research.com/articles/5-1499/v1
|
24 Jun 16
|
{
"type": "Review",
"title": "Statin-induced diabetes: incidence, mechanisms, and implications",
"authors": [
"Om P. Ganda"
],
"abstract": "Persuasive data from many randomized controlled trials and large, long-term observational studies indicate a modestly increased risk for the emergence of new diabetes after statin initiation. Several meta-analyses of many statin trials as well as longitudinal population-based studies suggest that the risk factors for diabetes in statin-treated persons include underlying risk for diabetes at baseline (specifically features of metabolic syndrome), the intensity of statin therapy, certain genetic traits independent of diabetes risk, and adherence to lifestyle factors. Limited data suggest statins modestly worsen hyperglycemia and A1c levels in those with pre-existing diabetes or glucose intolerance. The precise mechanism(s) of diabetogenesis with statin therapy are unclear, but impaired insulin sensitivity and compromised β cell function via enhanced intracellular cholesterol uptake due to inhibition of intracellular cholesterol synthesis by statins, as well as other mechanisms, may be involved. Furthermore, while statins are known to have anti-inflammatory effects, it is hypothesized that, under dysmetabolic conditions, they might have pro-inflammatory effects via induction of certain inflammasomes. This concept requires further elucidation in the human. Finally, it is clear that the risk–benefit ratio for cardiovascular disease events is strongly in favor of statin therapy in those at risk, despite the emergence of new diabetes. Adherence to lifestyle regimen is critical in the prevention of new diabetes on statins.",
"keywords": [
"diabetes",
"statin therapy",
"diabetogenesis"
],
"content": "Introduction\n\nThe remarkable value of HMG-CoA reductase inhibitors (statins) in atherosclerotic cardiovascular disease (CVD) risk reduction is clearly established, based on landmark secondary intervention as well as primary prevention trials during the past two decades1,2. The practical significance of a modest absolute risk reduction in primary prevention of CVD is still debated3,4. However, the recent long-term, global, multi-ethnic, primary prevention trial Heart Outcomes and Prevention Evaluation (HOPE)-3 has confirmed the benefits of a moderate dose of statin (rosuvastatin 10 mg) in subjects at an intermediate CVD risk, with a 24% reduction in primary CVD outcomes (hazard ratio [HR] 0.76; 95% confidence interval [CI] 0.66–0.88; p<0.001) over a mean follow-up of 5.6 years5. These results were consistent with those reported in the meta-analysis of 27 primary prevention statin trials2. While the total mortality or the CVD mortality in the 12,705 subjects in HOPE-3 was not reduced, a significant 15% reduction in CVD mortality and a significant 9% reduction in overall mortality per 1 mmol/L reduction in low-density lipoprotein cholesterol (LDL-C) over ~5 years were reported in the previous meta-analysis of the primary prevention trials2. Based on the overall evidence from the randomized controlled trials (RCTs), various updated major guidelines6–8 have paved the way for greater attention to initiation and intensification of statin therapy in high-risk individuals (such as those with prior CVD) and in individuals without CVD (such as those with diabetes and multiple risk factors).\n\n\nEmergence of new diabetes in RCTs\n\nA clinically relevant concern with statin therapy is a significantly increased risk of new-onset diabetes in patients on statin therapy. The JUPITER trial reported a 25% increase with rosuvastatin 20 mg, over a median follow-up of 1.9 years, compared to those on placebo9. Since then, several meta-analyses have confirmed a smaller but significant increase with various statins (Table 1). The analysis by Sattar et al. in 91,140 subjects showed a 9% overall risk in 13 RCTs over a mean period of 4.0 years (odds ratio [OR] 1.09; 95% CI 1.02–1.17)10. In a subsequent meta-analysis of five intensive-dose statin trials, Preiss et al. reported a significant increase in diabetes incidence with more intensive- vs. moderate-dose statin (OR 1.12; 95% CI 1.04–1.22) in 32,752 subjects over a mean follow-up of 4.9 years11. In general, there was no relationship between % LDL-C reduction and incident diabetes. Further analysis of baseline characteristics of the various trials reported a strong relationship between features of metabolic syndrome or pre-diabetes (age, body mass index [BMI], hypertension, fasting glucose, and triglycerides) at baseline and subsequent development of diabetes12–14.\n\nOf note, the risk–benefit ratio for CVD still clearly favored statin therapy in various studies, including JUPITER, in primary prevention13, several secondary prevention studies12,14, and a meta-analysis of secondary prevention studies by Preiss et al.11. Thus, regardless of whether or not diabetes was diagnosed during statin therapy, the CVD outcomes were reduced on statin therapy compared to those observed with placebo.\n\nAnother meta-analysis by Navarese et al. is the largest so far: it includes 17 RCTs (more than 113,000 patients). It compared new-onset diabetes in patients receiving statin vs. placebo, or high-dose vs. moderate-dose statins15. The lowest risk was seen with pravastatin 40 mg compared to placebo (OR 1.07; 95% CI 0.83–1.30), whereas rosuvastatin 20 mg was associated with the highest risk (OR 1.25; 95% CI 0.82–1.90) and atorvastatin 80 mg was intermediate (OR 1.15; 95% CI 0.9–1.50), even though none of these differences achieved statistical significance. Simvastatin also appears to be associated with higher risk compared to pravastatin. These differences among various statins persisted after adjustments for reduction in cholesterol. These findings suggest possible molecule-specific effects on diabetogenesis, although the data thus far are inconclusive.\n\nThe effects of the newest statin, pitavastatin, are not available in a large enough cohort. In a recent meta-analysis of 15 short-term RCTs of pitavastatin, most of 12 weeks’ duration, total follow-up 1600 person-years, there was no significant difference in the risk for diabetes (OR 0.70; 95% CI 0.30–1.61) compared to placebo16. If confirmed in a larger RCT, it will raise the possibility of differences in pharmacodynamics and drug-drug interactions on diabetogenecity.\n\n\nEmergence of new diabetes in population-based, observational studies\n\nTable 2 summarizes several large observational studies17–21 comparing patients on statins with those not on statins in various populations. These analyses revealed considerable variability among studies and with various statins, with HRs ranging from 1.19–1.57 but statistically significant, after follow-up durations of 3–6 years. In the Women’s Health study, the women were older than several other populations and generally on moderate-dose therapy, yet the HR was 1.4817. In the largest study of over 2 million subjects in the UK, there was a significant time-dependent increase in diabetes risk (HR 1.57; 95% CI 1.55–1.60), which increased further (HR 3.63; 95% CI 2.44–5.38) in those who were followed for up to 15–20 years21. In one study in patients following myocardial infarction, there was no difference in intensive- vs. moderate-dose statin therapy22, although the CVD outcomes were reduced with the more intensive approach. One caveat with all of the observational studies is that, despite multifactorial adjustments, some differences in the cohort characteristics may not be fully accounted for. In particular, it should be noted that the risk for diabetes according to presence of pre-existing diabetes risk factors, as observed in the several analyses of RCTs12–14, was not adequately examined in the various observational studies, a major limitation in those studies, compared to RCTs.\n\n\nGlucose control with statins in pre-existing diabetes or abnormal glucose tolerance\n\nThere are some observations of interest from a few studies in patients with pre-existing glucose intolerance or diabetes. In the study by Castro et al.19, the HR for progression to diabetes was similar in those with normoglycemia, or impaired fasting glucose at baseline, but both groups showed similar reduction in mortality after a 6-year follow-up. In a meta-analysis of nine RCTs in 9696 patients with type 2 diabetes, with a mean follow-up of 3.6 years, there was a modest but significant increase in mean A1c level of 0.12% (95% CI 0.04–0.20)23. In one cross-sectional study in patients with type 1 diabetes (n=1093), statin use was associated with a similar 0.2% increase in mean A1c after multivariate adjustments24.\n\n\nMechanisms underlying diabetogenic effects of statins\n\nThe precise mechanism(s) for statin-induced diabetes remain unclear, although the majority of patients developing diabetes have pre-diabetes or features of metabolic syndrome indicating high risk for diabetes at baseline12,13. It has been controversial whether chemical differences and pharmacodynamic differences in statins or more intensive statin therapy are more likely to precipitate diabetes. In the analysis by Preiss et al., intensive statin therapy led to a greater increase in diabetes11. This was also confirmed in other meta-analyses by Carter et al.25 and Dormuth et al.26. However, this was not confirmed in a propensity score-matched cohort of patients with myocardial infarction who were prescribed intensive- or moderate-dose statins and followed for 5 years (new diabetes in 13.6 vs. 13.0%)22. The reported lack of new diabetes in pitavastatin-treated subjects is intriguing in view of the relatively small and short-term studies with this newest statin so far (as discussed above)16. Another intriguing observation is that in the fairly large cohort of the HOPE-3 trial (n=12,705), there was no increase in the risk for new diabetes (HR 1.02; 95% CI 0.85–1.23) compared to a 25% increase with rosuvastatin 20 mg in JUPITER9. Whether this relates to the differences in the intensity of statin therapy, risk factors for diabetes at baseline, or perhaps genetic differences in the multi-ethnic HOPE-3 cohort is worth exploring.\n\nSeveral mechanisms have been postulated underlying the derangements in glucose metabolism by statins. There is some evidence for the detrimental effects of statins on both insulin sensitivity and β cell secretion. In the large METSIM observational study of more than 8000 men, simvastatin and atorvastatin were related to a dose-dependent increase in post-glucose load, an increase in glycemia, a mean decrease in insulin sensitivity by 24%, and a decline in insulin secretion by 12%18. Similarly, in a small study in 28 patients with polycystic ovary syndrome (PCOS), treatment with atorvastatin 20 mg compared to placebo over 6 months led to a decrease in insulin sensitivity despite a decrease in the inflammatory marker C-reactive protein (CRP)27.\n\nA number of potential deleterious effects of statins on β cell function have been proposed, including the effects of increased influx of cholesterol due to inhibition of HMG-CoA-mediated intracellular cholesterol synthesis, inhibition of ubiquinone (CoQ 10) synthesis leading to mitochondrial oxidative stress, and β cell apoptosis28. Statins are generally thought to have anti-inflammatory effects1,9. However, a recent novel hypothesis posits that under certain conditions, statins may activate inflammasome NLRP3 from macrophages or adipocytes in the presence of endotoxins, leading to interleukin-1β-mediated insulin resistance29. This hypothesis requires confirmation in human studies, as adipose tissue is not a major glucose-metabolizing tissue. A provocative possibility is that the altered gut microbiome, in the presence of obesity or other dysmetabolic states, might provide the endotoxin lipopolysaccharide (LPS) that may mediate the paradoxical pro-inflammatory effect of statins by activation of inflammasomes30. However, under physiological circumstances, a moderate decrease in insulin sensitivity should be compensated for by enhanced insulin secretion by β cells. Thus, ultimately, the direct or indirect effects of statins on β cell function may play an important role in diabetogenesis, particularly in those already at increased risk.\n\nAn exciting new observation is that a genetic polymorphism leading to a reduced activity of HMG-CoA reductase is associated with lower LDL-C, a significant increase in body weight, and features of insulin resistance31. This observation was validated in the randomized statin trials, and one particular allele was associated with a significant increase in the risk of new diabetes (OR 1.12; 95% CI 1.06–1.18). Since statins inhibit HMG-CoA reductase as their mode of action, this may at least partly explain their diabetogenic effect.\n\nFinally, a mundane and more simplistic nutritional explanation has been reported by the long-term data in the NHANES study. In a cross-sectional follow-up of more than 27,000 adults followed over 10 years, it was shown that those on statins liberalized their fat and total caloric intake and gained weight over time compared to those not on statins32. Thus, the progression to diabetes could be explained by lifestyle-induced worsening in insulin sensitivity. It is possible that this and the other postulated mechanisms could co-exist.\n\n\nImplications of statin-induced diabetes and worsening of hyperglycemia\n\nAs summarized above, there appears to be a significant relationship between statin use and the development of new diabetes over the course of several years. This has resulted in an understandable concern about the need to be more vigilant in the use of statins in primary prevention, particularly in those at low absolute risk of CVD. However, the outcome data, particularly from the RCTs, remind us that the subsequent events are also significantly reduced in those with statin-induced diabetes9–14. Sattar et al. calculated, based on their meta-analysis of 13 RCTs, that treatment with statins compared to placebo in 255 subjects over 4 years will cause one new case of diabetes while preventing 5.4 major CVD events10. Similarly, Preiss et al. estimated that intensive statin therapy, compared to moderate-dose statins, in those with prior CVD will prevent around three new events per year in ~500 subjects while resulting in one additional case of diabetes11. Finally, Ridker et al. reported that in the JUPITER trial, despite a 25% increase in the relative risk for new diabetes with rosuvastatin 20 mg, the major CVD event rate reduction in those who developed diabetes (HR 0.63; 95% CI 0.25–1.60) was consistent with event reductions in the trial as a whole (HR 0.56; 95% CI 0.46–0.69)13. Moreover, the risk of developing diabetes in that trial was almost entirely confined to those with pre-existing features of metabolic syndrome or pre-diabetes13, with similar data from other trials12. Also, in JUPITER, in those with risk factors for diabetes, 134 vascular events or deaths were avoided for every 54 new cases of new diabetes. Finally, in recent data from a cohort of more than 15,000 propensity-matched subjects who initiated statin therapy and were followed for a median of 2.7 years, there was no increase; in fact, there was a significant decrease in the development of microvascular complications (retinopathy and neuropathy) in those developing diabetes compared to a matched non-diabetic cohort33.\n\nThus, it seems quite clear that statin treatment should not be withheld in those at high risk of CVD for the relatively minor concern of progression to diabetes. In fact, the data described above indicate that in the presence of multiple risk factors for diabetes or metabolic syndrome at baseline in the at-risk population, a modest diabetogenic effect of statin therapy may lead to progression from pre-diabetes to diabetes. This should therefore prompt advice for lifestyle intervention, already known to prevent or delay progression to diabetes, and should be implemented prior to statin initiation. Moreover, the observations from the NHANES survey stated above32, that those on statin therapy generally increased their caloric intake and fat intake leading to progressive weight gain, i.e. factors known to be predictors of diabetes, further emphasize the need for lifestyle counseling as the integral component of both diabetes and CVD event reduction.",
"appendix": "Competing interests\n\n\n\nOm P. Ganda has received research support from Amarin Pharmaceuticals and consulting fees and speaker’s honoraria from Merck, Sanofi, and Amgen.\n\n\nGrant information\n\nOm P. Ganda’s work is supported by a NIH Diabetes Enrichment Core (P30DK36836).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe author thanks the reviewers for their helpful comments and suggestions.\n\n\nReferences\n\nCholesterol Treatment Trialists’ (CTT) Collaboration, Baigent C, Blackwell L, Emberson J, et al.: Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010; 376(9753): 1670–81. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCholesterol Treatment Trialists’ (CTT) Collaborators, Mihaylova B, Emberson J, Blackwell L, et al.: The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet. 2012; 380(984): 581–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTaylor F, Huffman MD, Macedo AF, et al.: Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2013; (1): CD004816. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nDuBroff R: Should Statin Therapy Be Guided by Cardiovascular Risk Models? Am J Med. 2016; 129(3): 235–7. PubMed Abstract | Publisher Full Text\n\nYusuf S, Bosch J, Dagenais G, et al.: Cholesterol Lowering in Intermediate-Risk Persons without Cardiovascular Disease. N Engl J Med. 2016; 374(21): 2021–31. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nStone NJ, Robinson JG, Lichtenstein AH, et al.: 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014; 63(25 Pt B): 2889–934. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nADA: Standards of Medical Care in Diabetes-2016: Summary of Revisions. Diabetes Care. 2016; 39(Suppl 1): S4–S5. PubMed Abstract | Publisher Full Text\n\nGanda OP: Deciphering cholesterol treatment guidelines: a clinician's perspective. JAMA. 2015; 313(10): 1009–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRidker PM, Danielson E, Fonseca FA, et al.: Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med. 2008; 359(21): 2195–207. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nSattar N, Preiss D, Murray HM, et al.: Statins and risk of incident diabetes: a collaborative meta-analysis of randomised statin trials. Lancet. 2010; 375(9716): 735–42. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPreiss D, Seshasai SR, Welsh P, et al.: Risk of incident diabetes with intensive-dose compared with moderate-dose statin therapy: a meta-analysis. JAMA. 2011; 305(24): 2556–64. PubMed Abstract | Publisher Full Text\n\nWaters DD, Ho JE, DeMicco DA, et al.: Predictors of new-onset diabetes in patients treated with atorvastatin: results from 3 large randomized clinical trials. J Am Coll Cardiol. 2011; 57(14): 1535–45. PubMed Abstract | Publisher Full Text\n\nRidker PM, Pradhan A, MacFadyen JG, et al.: Cardiovascular benefits and diabetes risks of statin therapy in primary prevention: an analysis from the JUPITER trial. Lancet. 2012; 380(9841): 565–71. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nWaters DD, Ho JE, Boekholdt SM, et al.: Cardiovascular event reduction versus new-onset diabetes during atorvastatin therapy: effect of baseline risk factors for diabetes. J Am Coll Cardiol. 2013; 61(2): 148–52. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nNavarese EP, Buffon A, Andreotti F, et al.: Meta-analysis of impact of different types and doses of statins on new-onset diabetes mellitus. Am J Cardiol. 2013; 111(8): 1123–30. PubMed Abstract | Publisher Full Text\n\nVallejo-Vaz AJ, Kondapally Seshasai SR, Kurogi K, et al.: Effect of pitavastatin on glucose, HbA1c and incident diabetes: A meta-analysis of randomized controlled clinical trials in individuals without diabetes. Atherosclerosis. 2015; 241(2): 409–18. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCulver AL, Ockene IS, Balasubramanian R, et al.: Statin use and risk of diabetes mellitus in postmenopausal women in the Women's Health Initiative. Arch Intern Med. 2012; 172(2): 144–52. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCederberg H, Stančáková A, Yaluri N, et al.: Increased risk of diabetes with statin treatment is associated with impaired insulin sensitivity and insulin secretion: a 6 year follow-up study of the METSIM cohort. Diabetologia. 2015; 58(5): 1109–17. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCastro MR, Simon G, Cha SS, et al.: Statin Use, Diabetes Incidence and Overall Mortality in Normoglycemic and Impaired Fasting Glucose Patients. J Gen Intern Med. 2016; 31(5): 502–8. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCorrao G, Ibrahim B, Nicotra F, et al.: Statins and the risk of diabetes: evidence from a large population-based cohort study. Diabetes Care. 2014; 37(8): 2225–32. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMacedo AF, Douglas I, Smeeth L, et al.: Statins and the risk of type 2 diabetes mellitus: cohort study using the UK clinical practice pesearch datalink. BMC Cardiovasc Disord. 2014; 14: 85. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nKo DT, Wijeysundera HC, Jackevicius CA, et al.: Diabetes mellitus and cardiovascular events in older patients with myocardial infarction prescribed intensive-dose and moderate-dose statins. Circ Cardiovasc Qual Outcomes. 2013; 6(3): 315–22. PubMed Abstract | Publisher Full Text\n\nErqou S, Lee CC, Adler AI: Statins and glycaemic control in individuals with diabetes: a systematic review and meta-analysis. Diabetologia. 2014; 57(12): 2444–52. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nJensen MT, Andersen HU, Rossing P, et al.: Statins are independently associated with increased HbA1c in type 1 diabetes--The Thousand & 1 Study. Diabetes Res Clin Pract. 2016; 111: 51–7. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCarter AA, Gomes T, Camacho X, et al.: Risk of incident diabetes among patients treated with statins: population based study. BMJ. 2013; 346: f2610. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nDormuth CR, Filion KB, Paterson JM, et al.: Higher potency statins and the risk of new diabetes: multicentre, observational study of administrative databases. BMJ. 2014; 348: g3244. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nPuurunen J, Piltonen T, Puukka K, et al.: Statin therapy worsens insulin sensitivity in women with polycystic ovary syndrome (PCOS): a prospective, randomized, double-blind, placebo-controlled study. J Clin Endocrinol Metab. 2013; 98(12): 4798–807. PubMed Abstract | Publisher Full Text\n\nSampson UK, Linton MF, Fazio S: Are statins diabetogenic? Curr Opin Cardiol. 2011; 26(4): 342–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenriksbo BD, Lau TC, Cavallari JF, et al.: Fluvastatin causes NLRP3 inflammasome-mediated adipose insulin resistance. Diabetes. 2014; 63(11): 3742–7. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMitchell P, Marette A: Statin-induced insulin resistance through inflammasome activation: sailing between Scylla and Charybdis. Diabetes. 2014; 63(11): 3569–71. PubMed Abstract | Publisher Full Text\n\nSwerdlow DI, Preiss D, Kuchenbaecker KB, et al.: HMG-coenzyme A reductase inhibition, type 2 diabetes, and bodyweight: evidence from genetic analysis and randomised trials. Lancet. 2015; 385(9965): 351–61. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nSugiyama T, Tsugawa Y, Tseng CH, et al.: Different time trends of caloric and fat intake between statin users and nonusers among US adults: gluttony in the time of statins? JAMA Intern Med. 2014; 174(7): 1038–45. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nNielsen SF, Nordestgaard BG: Statin use before diabetes diagnosis and risk of microvascular disease: a nationwide nested matched study. Lancet Diabetes Endocrinol. 2014; 2(11): 894–900. PubMed Abstract | Publisher Full Text | F1000 Recommendation"
}
|
[
{
"id": "14593",
"date": "24 Jun 2016",
"name": "Henry Ginsberg",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14594",
"date": "24 Jun 2016",
"name": "Michael H. Davidson",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1499
|
https://f1000research.com/articles/5-1498/v1
|
24 Jun 16
|
{
"type": "Review",
"title": "Models of intestinal infection by Salmonella enterica: introduction of a new neonate mouse model",
"authors": [
"Marc Schulte",
"Michael Hensel",
"Marc Schulte"
],
"abstract": "Salmonella enterica serovar Typhimurium is a foodborne pathogen causing inflammatory disease in the intestine following diarrhea and is responsible for thousands of deaths worldwide. Many in vitro investigations using cell culture models are available, but these do not represent the real natural environment present in the intestine of infected hosts. Several in vivo animal models have been used to study the host-pathogen interaction and to unravel the immune responses and cellular processes occurring during infection. An animal model for Salmonella-induced intestinal inflammation relies on the pretreatment of mice with streptomycin. This model is of great importance but still shows limitations to investigate the host-pathogen interaction in the small intestine in vivo. Here, we review the use of mouse models for Salmonella infections and focus on a new small animal model using 1-day-old neonate mice. The neonate model enables researchers to observe infection of both the small and large intestine, thereby offering perspectives for new experimental approaches, as well as to analyze the Salmonella-enterocyte interaction in the small intestine in vivo.",
"keywords": [
"Salmonella",
"intestinal inflammation",
"neonate mouse model",
"infection"
],
"content": "Introduction\n\nThe Gram-negative Enterobacteriaceae Salmonella enterica are among the main causes of bacterial gastrointestinal infections of millions of humans and animals around the world every year. The highest infection risk is oral ingestion of contaminated food or water often associated with insufficient hygiene conditions, but even industrial countries are not safe from infections1,2. About 2600 serovars of S. enterica are known, and serovars pathogenic to humans cause typhoidal and non-typhoidal forms of disease. The systemic disease enteric typhoid fever is caused by the human-restricted typhoidal serovars Typhi and Paratyphi A and is associated with high mortality if not treated by antibiotics. Non-typhoidal serovars, predominantly Enteritidis and Typhimurium, can infect a broad range of animals and humans, causing an acute self-limiting gastroenteritis associated with intestinal inflammation and diarrhea. This gastroenteritis is usually self-limiting without serious complications. However, systemic spread of non-typhoidal S. enterica may occur in infants, in the elderly, or in people with underlying infections or immunodeficiency (for example, due to HIV infection) and result in a more severe disease outcome. Persistent infections caused by non-typhoidal Salmonella are poorly investigated, and the prevalence of long-term non-typhoidal Salmonella carriers is not known3–6.\n\nSalmonella pathogenicity is mediated mainly by horizontally transferred chromosomal regions, encoding sets of virulence factors enabling the pathogen to successfully infect and colonize its host. The role in pathogenesis and the molecular functions of these so-called Salmonella pathogenicity islands (SPIs) are partly understood, but many functions remain to be resolved7. The SPI1 encodes a type III secretion system (T3SS) and the associated effector proteins, which can be injected into the host target cell (for example, epithelial cells) and thereby promote pathogen-induced internalization by non-phagocytic cells8–11.\n\nThe invasion of non-phagocytic cells by Salmonella and intracellular proliferation has been studied in much detail by using in vitro models12,13. Also, enterocyte invasion in different ex vivo tissue explants has been observed14–16. Nevertheless, these models bear limitations in studying the host-pathogen interaction in the natural anatomical environment, especially with respect to hallmarks of Salmonella pathogenesis – the invasion of polarized epithelial cells, intracellular survival, and formation of microcolonies.\n\nMurine infection models are attractive, since mice can be genetically manipulated. A previously developed mouse model for intestinal inflammation by S. enterica was based on antibiotic pretreatment of adult mice to reduce intestinal microbiota17. Here, we discuss a new infection model deploying 1-day-old neonate mice that allows the investigation of Salmonella enterocyte invasion, intracellular proliferation, and microcolony formation in vivo without pretreatment by antibiotics18.\n\n\nAnimal models of non-typhoidal Salmonella infections\n\nThe disease outcome of S. enterica infection, that is intestinal inflammation and diarrhea, or systemic infection with colonization of other organs is dependent on the host susceptibility and the serotype of the pathogen19. In humans or cattle, S. enterica serovar (sv.) Typhimurium induces enterocolitis, resulting in intestinal inflammation and diarrhea, whereas infected mice present no intestinal inflammation because of intrinsic resistance20. Nevertheless, certain mouse strains with defects in genes encoding SLC11A1 (previously named NRAMP1) develop a typhoid-like disease, similar to human infection with typhoidal serovars21. In infected susceptible mice, S. enterica sv. Typhimurium penetrates the epithelial barrier by invasion of microfold cells (M cells) or transport via dendritic cells22–24. M cells are specialized epithelial cells located in Peyer’s patches, which are organized lymphoid regions of the intestine. M cells phagocytose and transport antigens and bacteria to immune cells present in Peyer’s patches25. Salmonella-susceptible mice develop systemic infection after colonization of Peyer’s patches and mesenteric lymph nodes following spread to liver and spleen26,27. Owing to this pathogenesis and the absence of an appropriate small animal model, detailed analyses of Salmonella gastroenteritis were not possible. Earlier work investigated Salmonella-induced diarrhea in Rhesus monkeys28. As an alternative, infection of ligated murine and rabbit ileal loops was used as a model22,29. Furthermore, a bovine infection model was established that allowed the identification of certain Salmonella associated factors, such as the SPI1-T3SS, needed to induce enterocolitis20. In addition, bovine ligated ileal loops infected with Salmonella were investigated16,30,31. However, the use of large animals for infection causes technical limitations, and only restricted investigation of the role of the host in the host-pathogen interaction is possible. Hence, many features of Salmonella intestinal pathogenesis were analyzed by using tissue culture or intestinal organ culture32.\n\nOral application of the antibiotic streptomycin makes mice more susceptible to infection with Salmonella33–35. This effect was ascribed to removal of commensal intestinal microbes by streptomycin36,37. Based on these observations, a mouse model of oral infection of 6- to 8-week-old mice after pretreatment with streptomycin was established17, and this enabled the investigation of Salmonella-induced colitis in small animals. The pathogenesis of colitis caused by S. enterica sv. Typhimurium in streptomycin-pretreated mice showed many similarities to the human infection and its pathology is highly dependent on function of the SPI1-T3SS. With a knockout mouse strain that lacks all lymph nodes and organized gut-associated lymphatic tissues, it was shown that Peyer’s patches and mesenteric lymph nodes are not necessary for the induction of colitis17.\n\nHowever, there are some differences between the bovine and human infections and the infection in streptomycin-pretreated mice with regard to the intestinal location of the symptoms. Streptomycin-pretreated mice show inflammation of the cecum and the colon, whereas both the small and large intestine are affected in infections of cattle or humans20,38. Furthermore, the infection of rabbits, calves, and primates is often accompanied by massive luminal fluid secretion; however, streptomycin-pretreated mice do not show this phenomenon21. Translocation of Salmonella over the colonic epithelium in the absence of intracellular proliferation39 as well as enterocyte invasion and presence of Salmonella in the lamina propria in the mouse large intestine was demonstrated by using streptomycin-pretreated mice40–42. However, streptomycin-pretreated adult animals did not allow investigation of the infection process of the small intestinal epithelium, including invasion into polarized epithelial cells and intracellular survival.\n\nTo understand the role of bacterial as well as host factors for pathogenesis, it is of great importance to analyze the interaction of Salmonella with host cells within their natural environment. These factors, facilitating enterocyte invasion but also intraepithelial proliferation resulting in formation of intraepithelial bacterial colonies, remained undefined.\n\n\nA new neonate mouse model for Salmonella\n\nA small animal model using 1-day-old C57BL/6 mice was recently reported18 that may serve as an attractive alternative to the use of streptomycin-pretreated mice. An investigation of oral Salmonella infection of neonate and adult mice was accomplished, revealing age-dependent differences in intestinal colonization, mucosal translocation, and systemic spread. The work demonstrated rapid colonization of the small intestine and the colon of neonate mice, as opposed to adult animals as well as efficient entry of Salmonella into intestinal epithelial cells, followed by bacterial proliferation and formation of intraepithelial microcolonies.\n\nThe penetration of the mucosal barrier was dependent on enterocyte invasion and led to systemic spread to liver, spleen, and mesenteric lymph nodes. Without a functional SPI1-T3SS, systemic spread of Salmonella was largely abolished. Enterocytes were infected by wild-type, but not SPI1-T3SS-deficient, Salmonella. The major entry pathway for bacterial translocation is dependent on M cells23, but this cell population appears only after the neonatal period. It was shown that the expression of genes of differentiated M cells (for example, Spi-B and Ccl9) was very low in epithelial cells of neonate mice. Furthermore, no M cell markers like glycoprotein 2, Ulex europaeus agglutinin, or Ccl9 were found by immunostaining intestinal tissue. That shifts the major port of entry to enterocyte invasion and therefore is SPI1-T3SS dependent and M cell independent, whereas intestinal colonization and systemic spread in adult streptomycin-pretreated mice is SPI1 T3SS independent because of the presence of M cells. Moreover, the invasion of epithelial cells by Salmonella leads to intraepithelial proliferation and formation of intraepithelial microcolonies that arise from a single event of bacterial invasion. In addition, cells infected with Salmonella appear morphologically intact despite invasion and proliferation18. Differences and similarities of Salmonella infection models of neonate and adult mice are summarized in Figure 1.\n\nA comparison of 1-day-old, 6-day-old, and streptomycin-pretreated 6-week-old C57BL/6 mice shows many differences in intestinal colonization, mucosal translocation, and systemic dissemination in comparison with other organs after oral infection with Salmonella enterica serovar Typhimurium. Shading indicates no or low (red) and fully developed (green) features. M cell, microfold cell; SCV, Salmonella-containing vacuole.\n\nLow expression of mucin glycoproteins and lower thickness of the mucus layer were measured in neonate mice. Neonates showed a lower antimicrobial peptide repertoire in accordance with previous reports43. In contrast, adult animals exhibit the synthesis of various mucin glycoproteins, an enhanced thickness of the mucus layer, and antimicrobial peptides to create a strong barrier against pathogens44,45. A further difference between the epithelium of neonate and adult mice is an altered cyclin expression, resulting in a lower crypt to villus migration and minimal turnover of enterocytes, whereas enterocytes of adult animals show a constant renewal46. Therefore, the intestinal epithelial cells of neonate mice stay longer at the same position. It is speculated that the extended lifetime of epithelial cells and minimal expression of mucin glycoproteins and antimicrobial peptides as well as a reduced thickness of the mucus layer allow Salmonella to invade epithelial cells, to proliferate, and finally to form microcolonies. Another important difference is the lack of an established intestinal microbiota in neonate mice. A developed and diverse enteric microbiota is considered as a key factor of colonization resistance against Salmonella in adult mice. In the streptomycin pretreatment model, this colonization resistance is overcome by antibiotic elimination of a major fraction of the microbiota.\n\nAn innate immune response was further activated via Toll-like receptor (TLR) stimulation by intraepithelial, but not by non-invasive, extracellular Salmonella. After infection of neonate mice by wild-type Salmonella, a time-dependent increase in Cxcl2 and Cxcl5 mRNA expression was observed while mRNA expression was strongly reduced in enterocytes of adult mice isolated at day 4 after infection. This underlines the requirement of enterocyte invasion for the innate immune stimulation. Additionally, a large number of additional genes involved in metabolism, communication processes, and cellular responses were induced after infection. In the absence of innate immune receptors like TLR4, TLR2, TLR5, TLR9, MyD88, Unc93B1, and Nod2, the expression of Cxcl2 and Reg3γ mRNA by epithelial cells was severely reduced. Even in the absence of the most effective innate immune receptor, TLR4, an intestinal colonization of Salmonella as well as enterocyte invasion and spread to systemic organs were observed. Compared with infection of adult hosts, neonatal intestinal tissue remained largely intact in terms of histopathological parameters and epithelial barrier integrity.\n\nThere are some important similarities such as the constituents of the mature Salmonella-containing vacuole as well as autophagosomal factors that are expressed by both neonatal and adult epithelial cells. This may enable the future investigation of the intracellular lifestyle of Salmonella during the initial phase of infection.\n\n\nAdvantages, limitations, and future perspectives\n\nIn contrast to other infection models (for instance, the bovine host), the mouse is amenable to efficient genetic manipulation and therefore offers the opportunity to analyze the host-pathogen interaction with both genetically altered pathogen and host (Table 1). Owing to the absence of M cells in the neonate host, Salmonella-enterocyte interaction as well as invasion of polarized epithelial cells, intraepithelial proliferation, and the formation of microcolonies can be observed in their natural anatomical environment, which is a requirement for understanding the contribution of bacterial virulence factors. This may enable researchers to characterize the early steps in Salmonella pathogenesis and to discover the functional role of effector proteins injected into the host cell via secretion systems. The analysis of SPI1-T3SS and SPI2-T3SS effector translocation can perhaps be followed by live cell imaging by using effector proteins fused to self-labeling enzyme tags. Moreover, new approaches to investigate the innate immune response of the host are made possible. The morphological features of cells infected with Salmonella are largely unaltered and, despite an innate immune response, the severe damage of the epithelium observed in adult animals is absent in infected neonates.\n\nThe neonate mouse animal model provides advantages in investigating the Salmonella-enterocyte interaction in vivo, but there are some limitations like the small animal size and a lack of suitable anesthesia needed for intravital microscopy that have to be considered. Owing to the lack of M cells, one important route of infection is not represented in the neonate model. The role of intestinal microbiota during Salmonella infection cannot be addressed, since the microbiome in neonates is highly reduced and distinct from the microbiota of adult individuals. Regardless, the new neonate animal model could contribute to a better understanding of the cellular processes during infection. The model allows in vivo analyses of both hallmarks of Salmonella pathogenesis: the internalization by non-phagocytic cells and the intracellular activity of Salmonella leading to the formation of intraepithelial microcolonies.\n\n\nAbbreviations\n\nM cell, microfold cell; SPI, Salmonella pathogenicity island; T3SS, type 3 secretion system; TLR, Toll-like receptor.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nWork in the group of Michael Hensel was funded by the DFG, the BMBF, and the MWK Niedersachsen.\n\n\nAcknowledgements\n\nWe thank Mathias Hornef, Uniklinik RWTH Aachen, for sharing unpublished data.\n\n\nReferences\n\nBrenner FW, Villar RG, Angulo FJ, et al.: Salmonella nomenclature. J Clin Microbiol. 2000; 38(7): 2465–7. PubMed Abstract | Free Full Text\n\nHaraga A, Ohlson MB, Miller SI: Salmonellae interplay with host cells. Nat Rev Microbiol. 2008; 6(1): 53–66. PubMed Abstract | Publisher Full Text\n\nGunn JS, Marshall JM, Baker S, et al.: Salmonella chronic carriage: epidemiology, diagnosis, and gallbladder persistence. Trends Microbiol. 2014; 22(11): 648–55. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMarzel A, Desai PT, Goren A, et al.: Persistent Infections by Nontyphoidal Salmonella in Humans: Epidemiology and Genetics. Clin Infect Dis. 2016; 62(7): 879–86. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGal-Mor O, Boyle EC, Grassl GA: Same species, different diseases: how and why typhoidal and non-typhoidal Salmonella enterica serovars differ. Front Microbiol. 2014; 5: 391. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSubramoney EL: Non-typhoidal Salmonella infections in HIV-positive adults. S Afr Med J. 2015; 105(10): 805–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGerlach RG, Hensel M: Salmonella pathogenicity islands in host specificity, host pathogen-interactions and antibiotics resistance of Salmonella enterica. Berl Munch Tierarztl Wochenschr. 2007; 120(7-8): 317–27. PubMed Abstract\n\nQue F, Wu S, Huang R: Salmonella pathogenicity island 1(SPI-1) at work. Curr Microbiol. 2013; 66(6): 582–7. PubMed Abstract | Publisher Full Text\n\nSchlumberger MC, Hardt WD: Salmonella type III secretion effectors: pulling the host cell's strings. Curr Opin Microbiol. 2006; 9(1): 46–54. PubMed Abstract | Publisher Full Text\n\nGerlach RG, Hensel M: Protein secretion systems and adhesins: the molecular armory of Gram-negative pathogens. Int J Med Microbiol. 2007; 297(6): 401–15. PubMed Abstract | Publisher Full Text\n\nCollazo CM, Galán JE: The invasion-associated type-III protein secretion system in Salmonella--a review. Gene. 1997; 192(1): 51–9. PubMed Abstract | Publisher Full Text\n\nFigueira R, Watson KG, Holden DW, et al.: Identification of Salmonella pathogenicity island-2 type III secretion system effectors involved in intramacrophage replication of S. enterica serovar typhimurium: implications for rational vaccine design. MBio. 2013; 4(2): e00065. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRappl C, Deiwick J, Hensel M: Acidic pH is required for the functional assembly of the type III secretion system encoded by Salmonella pathogenicity island 2. FEMS Microbiol Lett. 2003; 226(2): 363–72. PubMed Abstract | Publisher Full Text\n\nNietfeld JC, Tyler DE, Harrison LR, et al.: Invasion of enterocytes in cultured porcine small intestinal mucosal explants by Salmonella choleraesuis. Am J Vet Res. 1992; 53(9): 1493–9. PubMed Abstract\n\nGiannella RA, Formal SB, Dammin GJ, et al.: Pathogenesis of salmonellosis. Studies of fluid secretion, mucosal invasion, and morphologic reaction in the rabbit ileum. J Clin Invest. 1973; 52(2): 441–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrost AJ, Bland AP, Wallis TS: The early dynamic response of the calf ileal epithelium to Salmonella typhimurium. Vet Pathol. 1997; 34(5): 369–86. PubMed Abstract | Publisher Full Text\n\nBarthel M, Hapfelmeier S, Quintanilla-Martínez L, et al.: Pretreatment of mice with streptomycin provides a Salmonella enterica serovar Typhimurium colitis model that allows analysis of both pathogen and host. Infect Immun. 2003; 71(5): 2839–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang K, Dupont A, Torow N, et al.: Age-dependent enterocyte invasion and microcolony formation by Salmonella. PLoS Pathog. 2014; 10(9): e1004385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUzzau S, Brown DJ, Wallis T, et al.: Host adapted serotypes of Salmonella enterica. Epidemiol Infect. 2000; 125(2): 229–55. PubMed Abstract | Free Full Text\n\nSantos RL, Zhang S, Tsolis RM, et al.: Animal models of Salmonella infections: enteritis versus typhoid fever. Microbes Infect. 2001; 3(14–15): 1335–44. PubMed Abstract | Publisher Full Text\n\nTsolis RM, Kingsley RA, Townsend SM, et al.: Of mice, calves, and men. Comparison of the mouse typhoid model with other Salmonella infections. Adv Exp Med Biol. 1999; 473: 261–74. PubMed Abstract | Publisher Full Text\n\nClark MA, Jepson MA, Simmons NL, et al.: Preferential interaction of Salmonella typhimurium with mouse Peyer's patch M cells. Res Microbiol. 1994; 145(7): 543–52. PubMed Abstract | Publisher Full Text\n\nJones BD, Ghori N, Falkow S: Salmonella typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer's patches. J Exp Med. 1994; 180(1): 15–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVazquez-Torres A, Jones-Carson J, Bäumler AJ, et al.: Extraintestinal dissemination of Salmonella by CD18-expressing phagocytes. Nature. 1999; 401(6755): 804–8. PubMed Abstract | Publisher Full Text\n\nJang MH, Kweon MN, Iwatani K, et al.: Intestinal villous M cells: an antigen entry site in the mucosal epithelium. Proc Natl Acad Sci U S A. 2004; 101(16): 6110–5. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nCarter PB, Collins FM: The route of enteric infection in normal mice. J Exp Med. 1974; 139(5): 1189–203. PubMed Abstract | Free Full Text\n\nHohmann AW, Schmidt G, Rowley D: Intestinal colonization and virulence of Salmonella in mice. Infect Immun. 1978; 22(3): 763–70. PubMed Abstract | Free Full Text\n\nRout WR, Formal SB, Dammin GJ, et al.: Pathophysiology of Salmonella diarrhea in the Rhesus monkey: Intestinal transport, morphological and bacteriological studies. Gastroenterology. 1974; 67(1): 59–70. PubMed Abstract\n\nPenheiter KL, Mathur N, Giles D, et al.: Non-invasive Salmonella typhimurium mutants are avirulent because of an inability to enter and destroy M cells of ileal Peyer's patches. Mol Microbiol. 1997; 24(4): 697–709. PubMed Abstract | Publisher Full Text\n\nBolton AJ, Osborne MP, Wallis TS, et al.: Interaction of Salmonella choleraesuis, Salmonella dublin and Salmonella typhimurium with porcine and bovine terminal ileum in vivo. Microbiology. 1999; 145(Pt 9): 2431–41. PubMed Abstract | Publisher Full Text\n\nSantos RL, Tsolis RM, Zhang S, et al.: Salmonella-induced cell death is not required for enteritis in calves. Infect Immun. 2001; 69(7): 4610–7. PubMed Abstract | Publisher Full Text\n\nGalán JE: Salmonella interactions with host cells: type III secretion at work. Annu Rev Cell Dev Biol. 2001; 17: 53–86. PubMed Abstract | Publisher Full Text\n\nBohnhoff M, Drake BL, Miller CP: Effect of streptomycin on susceptibility of intestinal tract to experimental Salmonella infection. Proc Soc Exp Biol Med. 1954; 86(1): 132–7. PubMed Abstract | Publisher Full Text\n\nBohnhoff M, Miller CP: Enhanced susceptibility to Salmonella infection in streptomycin-treated mice. J Infect Dis. 1962; 111: 117–27. PubMed Abstract | Publisher Full Text\n\nMeynell GG, Subbaiah TV: Antibacterial mechanisms of the mouse gut. I. Kinetics of infection by Salmonella typhi-murium in normal and streptomycin-treated mice studied with abortive transductants. Br J Exp Pathol. 1963; 44: 197–208. PubMed Abstract | Free Full Text\n\nBohnhoff M, Miller CP, Martin WR: Resistance of the mouse's intestinal tract to experimental salmonella infection. II. Factors responsible for its loss following streptomycin treatment. J Exp Med. 1964; 120: 817–28. PubMed Abstract | Free Full Text\n\nBohnhoff M, Miller CP, Martin WR: Resistance of the mouse's intestinal tract to experimental salmonella infection. I. Factors which interfere with the initiation of infection by oral inoculation. J Exp Med. 1964; 120: 805–16. PubMed Abstract | Free Full Text\n\nTsolis RM, Townsend SM, Miao EA, et al.: Identification of a putative Salmonella enterica serotype typhimurium host range factor with homology to IpaH and YopM by signature-tagged mutagenesis. Infect Immun. 1999; 67(12): 6385–93. PubMed Abstract | Free Full Text\n\nMüller AJ, Kaiser P, Dittmar KE, et al.: Salmonella gut invasion involves TTSS-2-dependent epithelial traversal, basolateral exit, and uptake by epithelium-sampling lamina propria phagocytes. Cell Host Microbe. 2012; 11(1): 19–32. PubMed Abstract | Publisher Full Text\n\nHapfelmeier S, Stecher B, Barthel M, et al.: The Salmonella pathogenicity island (SPI)-2 and SPI-1 type III secretion systems allow Salmonella serovar typhimurium to trigger colitis via MyD88-dependent and MyD88-independent mechanisms. J Immunol. 2005; 174(3): 1675–85. PubMed Abstract | Publisher Full Text\n\nFelmy B, Songhet P, Slack EM, et al.: NADPH oxidase deficient mice develop colitis and bacteremia upon infection with normally avirulent, TTSS-1- and TTSS-2-deficient Salmonella Typhimurium. PLoS One. 2013; 8(10): e77204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMüller AJ, Hoffmann C, Galle M, et al.: The S. Typhimurium effector SopE induces caspase-1 activation in stromal cells to initiate gut inflammation. Cell Host Microbe. 2009; 6(2): 125–36. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMénard S, Förster V, Lotz M, et al.: Developmental switch of intestinal antimicrobial peptide expression. J Exp Med. 2008; 205(1): 183–93. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nZarepour M, Bhullar K, Montero M, et al.: The mucin Muc2 limits pathogen burdens and epithelial barrier dysfunction during Salmonella enterica serovar Typhimurium colitis. Infect Immun. 2013; 81(10): 3672–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDupont A, Kaconis Y, Yang I, et al.: Intestinal mucus affinity and biological activity of an orally administered antibacterial and anti-inflammatory peptide. Gut. 2015; 64(2): 222–32. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nde Santa Barbara P, van den Brink GR, Roberts DJ: Development and differentiation of the intestinal epithelium. Cell Mol Life Sci. 2003; 60(7): 1322–32. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "14589",
"date": "24 Jun 2016",
"name": "Ohad Gal-Mor",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14590",
"date": "24 Jun 2016",
"name": "John S Gunn",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1498
|
https://f1000research.com/articles/5-1497/v1
|
24 Jun 16
|
{
"type": "Review",
"title": "Recent advances in understanding ichthyosis pathogenesis",
"authors": [
"Nareh V. Marukian",
"Keith A. Choate",
"Nareh V. Marukian"
],
"abstract": "The ichthyoses, also known as disorders of keratinization (DOK), encompass a heterogeneous group of skin diseases linked by the common finding of abnormal barrier function, which initiates a default compensatory pathway of hyperproliferation, resulting in the characteristic clinical manifestation of localized and/or generalized scaling. Additional cutaneous findings frequently seen in ichthyoses include generalized xerosis, erythroderma, palmoplantar keratoderma, hypohydrosis, and recurrent infections. In 2009, the Ichthyosis Consensus Conference established a classification consensus for DOK based on pathophysiology, clinical manifestations, and mode of inheritance. This nomenclature system divides DOK into two main groups: nonsyndromic forms, with clinical findings limited to the skin, and syndromic forms, with involvement of additional organ systems. Advances in next-generation sequencing technology have allowed for more rapid and cost-effective genetic analysis, leading to the identification of novel, rare mutations that cause DOK, many of which represent phenotypic expansion. This review focuses on new findings in syndromic and nonsyndromic ichthyoses, with emphasis on novel genetic discoveries that provide insight into disease pathogenesis.",
"keywords": [
"ichthyosis",
"keratinization",
"hyperproliferation",
"pathogenesis"
],
"content": "Introduction\n\nThe ichthyoses encompass a heterogeneous group of skin diseases linked by the common finding of abnormal barrier function, which leads to increased transepidermal water loss and compensatory hyperproliferation. The unifying phenotypic feature of the ichthyoses is localized and/or generalized scaling. Other clinical manifestations can include erythroderma (confluent red skin), palmoplantar keratoderma (thickening of the palms and soles), hypohydrosis (diminished sweating), and recurrent infections.\n\nAlthough ichthyoses are primarily inherited disorders with onset at or shortly after birth, rare acquired forms have been reported in the setting of malignancy, nutritional deficiency, and autoimmune or infectious disease. Mutations in over 50 genes have been reported to cause ichthyoses, and these affect a host of cellular functions including DNA repair, lipid biosynthesis, adhesion, and desquamation as well as other pathways1. Despite myriad pathways for pathogenesis, each features disrupted barrier function.\n\nEpidermal barrier function is maintained by a regular pattern of epidermal renewal in which keratinocytes, the primary cell type of the skin, arise from a renewing stem cell pool and undergo a tightly regulated pattern of differentiation as they transit from the innermost stratum basale to the outermost stratum corneum, where they are ultimately sloughed off. This differentiation program is marked by site-specific expression of proteins and, in the suprabasal layers, the production of components necessary for the generation of the lipid barrier.\n\nIn the process of differentiation, keratins—intermediate filaments that are responsible for the structural integrity of keratinocytes—are among the first proteins to be expressed in a tightly regulated manner, with keratin 5 and 14 expressed in the basal layer and keratin 1 and 10 expressed in the suprabasal layers. In the stratum spinosum, the second innermost layer of the epidermis, components of the lipid barrier (phospholipids, cholesterol, sphingomyelin, and glucosylceramides) are packaged into lamellar bodies, which are specialized organelles that house the building blocks of the lipid barrier as well as enzymes essential to the processing of lipid barrier precursors. At the transition from the stratum granulosum—the third layer of the epidermis—to the stratum corneum, the contents of lamellar bodies are extruded into the intercellular space to form protective lipid sheets that are responsible for the skin’s hydrophobic barrier2,3 (Figure 1).\n\nEpidermal barrier function is maintained by a regular pattern of epidermal differentiation and generation of lipid components.\n\nThis overall process of differentiation results in the formation of a robust barrier in the stratum corneum, composed of keratinocytes (the individual bricks of the barrier) and inter-keratinocyte lipids (the mortar)4. Mutations in proteins essential to the formation of this barrier (i.e. keratins and enzymes involved in lipid synthesis) lead to the disruption of barrier integrity, resulting in ichthyosis.\n\nInherited ichthyoses exhibit marked genetic and phenotypic heterogeneity, and advances in next-generation sequencing technology have allowed for more rapid and cost-effective genetic analysis, leading to the identification of novel, rare mutations that cause DOK. Clear large-scale genotype-phenotype correlations have been difficult to establish, as mutations in the same gene can present with widely divergent phenotypes, even within kindreds bearing the same disease-causing mutation.\n\nIn 2009, the Ichthyosis Consensus Conference established a consensus classification for DOK based on pathophysiology, clinical manifestations, and mode of inheritance1. This nomenclature system divides DOK into two main groups: 1) nonsyndromic forms, with clinical findings limited to the skin, and 2) syndromic forms, with involvement of other organ systems.\n\n\nNonsyndromic ichthyoses\n\nIchthyosis vulgaris (IV) and X-linked recessive ichthyosis (XLRI) are classified as the “common ichthyoses”, given their high prevalence. IV is the most common form of nonsyndromic inherited ichthyosis, with an estimated incidence of 1 in 250 births5. Typically, IV is a phenotypically mild form of ichthyosis. Clinical findings usually appear at around 2 months of age and include generalized xerosis and fine white to gray scale that is most prominent on the abdomen, chest, and extensor surfaces of the extremities. Keratosis pilaris and hyper-linearity of the palms and soles are also frequently associated with IV.\n\nIV is caused by autosomal dominant mutations in the filaggrin gene (FLG), which plays an essential role in epidermal differentiation and formation of the skin barrier6,7. An autosomal semidominant mode of inheritance has also been described, meaning that while individuals with heterozygous mutations have a mild phenotype, those with homozygous or compound heterozygous mutations can display more severe forms of ichthyosis6.\n\nPatients with IV are at increased risk for atopic dermatitis, asthma, and allergies8,9. This increased risk is likely due to disruption of barrier function, which may allow for greater penetration of the epidermis by potential allergens8.\n\nXLRI is the second most common form of inherited ichthyosis, with a prevalence of 1:2000 to 1:6000 in males10. Clinical findings in XLRI are frequently indistinguishable from IV. Manifestations usually first appear in the neonatal period as generalized desquamation and xerosis and progress to fine scaling of the trunk and extremities in infancy. Over time, patients develop brownish, polygonal, plate-like scale that is tightly adherent to the skin. XLRI is caused by mutations in the STS gene, encoding steroid sulfatase, on the X chromosome11.\n\nAutosomal recessive congenital ichthyosis (ARCI) is a genetically and phenotypically heterogeneous group of disorders that includes harlequin ichthyosis (HI), lamellar ichthyosis (LI), and congenital ichthyosiform erythroderma (CIE). The incidence of ARCI has been approximated at 1 in 200,000 births12.\n\nHI is caused by loss-of-function mutations in ABCA12, which encodes an ATP-binding cassette (ABC) transporter. ABCA12 is necessary for lipid transport into lamellar granules and is central to the process of cornification and lipid barrier formation13. Interestingly, while homozygous loss-of-function mutations in ABCA12 lead to HI, missense mutations in ABCA12 result in milder phenotypes on the LI/CIE spectrum14. Neonates with HI present with thick, armor-like scale with severe ectropion (eversion of the eyelids), eclabium (eversion of the lips), and flattening of the ears. Some patients with HI die during the neonatal period, but survival has been shown to improve with progress in neonatal intensive care and early treatment with systemic retinoids. Rajpopat et al. showed that 83% of HI patients treated with oral retinoids survived compared to 24% of untreated patients15.\n\nLI and CIE represent a spectrum of disorders caused by mutations in one of nine genes: TGM1, NIPAL4/ICHTHYIN, ALOX12B, ALOXE3, CYP4F22, ABCA12, PNPLA1, CERS3, and LIPN16. Mutations in TGM1 are the most common and account for approximately 32% of heritability of ARCI17. Fisher et al. found that mutations in the six most common genes (TGM1, NIPAL4, ALOX12B, CYP4F22, ALOXE3, and ABCA12) account for 78% of ARCI cases17. Despite this, prior studies of large cohorts of patients with ARCI showed that 22-40% of patients have no mutations in known genes17,18, highlighting the heterogeneity of this group of disorders and the importance of continued efforts in gene discovery.\n\nKeratinopathic ichthyosis is a group of disorders caused by mutations in the keratin family of genes. The major variant of keratinopathic ichthyosis is epidermolytic ichthyosis (EI). Minor variants include superficial EI (SEI), annular EI (AEI), and ichthyosis Curth-Macklin.\n\nEI is caused by autosomal dominant mutations in the keratin 1 (KRT1) and keratin 10 (KRT10) genes, which play an essential role in maintaining structural integrity in suprabasal keratinocytes19. EI is characterized by marked skin fragility, leading to generalized blister formation on a background of erythroderma. Neonates present with blistering and erythema at birth, but symptoms improve over time. Blistering becomes less frequent and is usually confined to sites of trauma in adulthood. Palmoplantar keratoderma is often associated with EI, although it is more commonly seen in patients with mutations in KRT1 than KRT1019.\n\nSEI, also known as ichthyosis bullosa of Siemens, is caused by mutations in KRT220,21. Phenotypic manifestations are milder compared to EI and include blister formation in response to trauma and hyperkeratosis (thickening of the stratum corneum) over flexural areas.\n\nAEI is a rare phenotypic variant of EI that was shown by Yang et al. to be caused by a unique mutation in KRT10 that replaces an alanine at residue 12 with a proline22. It is characterized by blister formation at birth, which later progresses to the intermittent development of annular polycyclic erythematous plaques on the trunk and extremities.\n\nIchthyosis Curth-Macklin is another rare disorder and is caused by autosomal dominant mutations in KRT123,24. It is characterized by extensive spiky or verrucous hyperkeratosis over the trunk and extensor surfaces of the extremities. It may also be associated with severe palmoplantar keratoderma. In the past 5 years, two novel distinct causative mutations in KRT1 have been identified in addition to the two mutations that were initially described25,26. While both EI and ichthyosis Curth-Macklin can be caused by mutations in KRT1, EI is caused by amino acid substitutions and in-frame deletions in the gene27, while ichthyosis Curth-Macklin is caused by insertions or deletions that lead to a frameshift23–26.\n\n\nSyndromic ichthyoses\n\nIn addition to cutaneous involvement, syndromic ichthyoses affect at least one other organ or system. Many causative genes have been identified for syndromic ichthyoses, including NSDHL (CHILD syndrome)28, EBP (Conradi-Hunermann-Happle syndrome, CHILD Syndrome)28,29, and ALDH3A2 (Sjögren-Larsson syndrome)30. Depending on the specific gene mutated, a wide range of organ systems can be involved, including the skeletal, nervous, endocrine, and cardiovascular systems. Many of the syndromic ichthyoses may present at birth with isolated cutaneous findings, highlighting the importance of a high degree of clinical suspicion and the usefulness of genetic analysis in the early diagnosis of these syndromic cases.\n\n\nRecent advances in ichthyosis\n\nMutations in PNPLA1 cause autosomal recessive congenital ichthyosis. In 2012, Grall et al. reported that mutations in the patatin-like phospholipase domain-containing protein 1 gene (PNPLA1) cause ARCI in Golden Retriever dogs and humans31. Selective inbreeding of dogs to create pure breeds leads to the propagation of not only specific desirable traits but also disease-causing alleles. The selection of these undesirable alleles results in the high prevalence of breed-specific diseases in dogs. For example, the inbreeding of Golden Retrievers has led to the high prevalence of ichthyosis within the breed. The frequency of the mutation in Golden Retrievers is estimated to be approximately 50%31. Ichthyosis in Golden Retrievers presents with generalized scaling, with white or black scale, similar to the phenotypic manifestations of ichthyosis in humans.\n\nIntermarriage within families and breeding approaches for purebred animals provide a unique opportunity to study the genetic basis of rare conditions. Grall et al. performed genetic analysis on 20 affected Golden Retrievers, which revealed homozygous mutations in PNPLA1 in all members of the cohort. Further studies on a human cohort of 46 consanguineous families with ARCI, who were previously found not to have mutations in known ARCI genes, revealed two distinct mutations in PNPLA1 in two different families31.\n\nThe PNPLA family of proteins contains nine members, which play key roles in lipid metabolism32,33. While disease-causing mutations in other members of the family had been previously identified, mutations in PNPLA1 had not been previously implicated in any disease34–37. This finding not only expands the genetic understanding of ARCI but also highlights the essential role of PNPLA1 in lipid metabolism and maintenance of the barrier function.\n\nMutations in GJA1 cause erythrokeratodermia variabilis et progressiva. In 2015, Boyden et al. reported that autosomal dominant mutations in GJA1 cause erythrokeratodermia variabilis et progressiva (EKVP)38, a rare genetic disorder characterized by transient figurate erythematous patches on a background of generalized scaling. GJA1 encodes connexin 43 (Cx43), which is present throughout the epidermis and is expressed in every tissue type39. Connexins, also known as gap junction proteins, are classified into alpha and beta subgroups40, encoded by GJA and GJB genes, respectively. Individual connexins form hexamers called connexons.\n\nConnexons can serve two main functions within the plasma membrane—individual connexons can either function as hemichannels, allowing for communication between the cytoplasm and the extracellular space, or dock with connexons on neighboring cells to form gap junctions. Gap junctions are essential to intercellular communication, allowing for synchronization of metabolic and electrical activities between cells as well as the exchange of small molecules and ions. Mutations in connexin genes have been previously shown to cause a wide range of disease phenotypes, including myelin-related disease, skin disease, hearing loss, and congenital cataract41.\n\nWhile mutations in GJB3 and GJB4 have been previously shown to cause EKV/EKVP42,43, Boyden et al. were first to report that mutations in GJA1 can also cause the phenotype38. Mutations in GJA1 have been previously described to cause oculodentodigital dysplasia (ODDD)39,44, which is a systemic disorder with limited cutaneous findings and sharply contrasts with the widespread cutaneous findings with lack of systemic symptoms seen in EKVP. Mutations in GJA1 that result in EKVP lead to mislocalization of Cx4338, while mutations resulting in ODDD lead to functionally impaired gap junctions that show a normal pattern of localization45. This finding expands the genetic understanding of EKVP and provides insight into its molecular mechanism.\n\nMutations in CARD14 cause pityriasis rubra pilaris. Pityriasis rubra pilaris (PRP) is a papulosquamous disorder that is characterized by well-demarcated salmon-colored plaques with fine scale, palmoplantar keratoderma, and follicular hyperkeratosis (excessive accumulation of keratin in hair follicles) that presents shortly after birth or can be acquired later in life, typically in the fourth or fifth decade. Although phenotypic features of PRP overlap with psoriasis, the two can be distinguished based on distinct clinical and histopathological factors46–49. There is much debate over the pathogenesis of PRP: infectious, inflammatory, and vitamin A-associated etiologies have been proposed46,50–52. A small fraction of PRP cases (less than 5%) are familial and are inherited in an autosomal dominant fashion46,53,54.\n\nIn 2012, Fuchs-Telem et al. studied four unrelated families with familial PRP and identified three distinct mutations in caspase recruitment domain family member 14 (CARD14)55. CARD14 is a known modulator of nuclear factor kappa B (NF-κB), which plays an important role in inflammatory pathways56,57. Fuchs-Telem et al. showed that NF-κB signaling is activated in PRP-affected skin and suggested that this inflammatory upregulation may play a role in the pathogenesis of familial PRP. Interestingly, causative mutations in CARD14 were previously identified in familial psoriasis, and enhanced NF-κB signaling was also identified as a possible pathogenic factor58. Taken together, these findings indicate that in addition to similarities in phenotypic features, familial PRP and familial psoriasis may share a common pathophysiology. Given the overall poor response to current treatments for PRP, this finding may allow for new therapeutic approaches that are aimed at modulating the immune response.\n\nSpecific mutations in TGM1 cause bathing-suit ichthyosis. As discussed above, ARCI encompasses a wide range of phenotypes, including LI, CIE, and HI. The most common underlying gene defect is the tranglutaminase-1 gene (TGM1), which accounts for approximately 30% of heritability of ARCI17.\n\nBathing-suit ichthyosis (BSI) is a rare variant of ARCI and is distinguished from the other forms of ARCI by restriction of scaling to the trunk, with sparing of the central face, buttocks, and limbs. All cases of BSI published to date have been caused by mutations in TGM1, although mutations in TGM1 more commonly cause generalized ARCI59–61. In 2006, Oji et al. showed that the specific TGM1 mutations that cause BSI may lead to temperature-dependent activity of transglutaminase, with marked decrease in enzyme function at higher temperatures61. This may account for the differential scaling observed in BSI, with greater disease manifestation at sites with relatively higher temperature, such as the trunk. More recently, several studies have identified additional mutations in TGM1 that contribute to BSI62–64, furthering the genetic understanding of this rare type of ichthyosis.\n\nMutations in DSP cause erythrokeratodermia-cardiomyopathy syndrome. In 2016, Boyden et al. identified a novel cardio-cutaneous disorder known as erythrokeratodermia-cardiomyopathy (EKC) syndrome. Early clinical findings include generalized erythrokeratodermia, recurrent infections, and failure to thrive as well as wiry or absent hair, dental enamel defects, absence of secondary teeth, and nail dystrophy. A hallmark of EKC is initially asymptomatic, rapidly progressive, potentially fatal cardiomyopathy, which was found in all three patients with EKC published in the literature to date65.\n\nEKC is caused by mutations in DSP, which encodes desmoplakin, a primary component of desmosomes65. Desmosomes are intercellular adhesion junctions that are most abundant in the epidermis and the heart. DSP has been previously implicated in several disorders, including diseases with isolated cardiac manifestations and cardio-cutaneous syndromes. Examples of disorders caused by mutations in DSP include striate palmoplantar keratoderma, smooth palmoplantar keratoderma with woolly hair, Carvajal syndrome (dilated cardiomyopathy with woolly hair and keratoderma), and arrhythmogenic cardiomyopathy66–70. However, EKC syndrome represents a distinct clinical phenotype and is the only disorder caused by mutations in DSP to present with erythrokeratodermia65.\n\nAlthough EKC syndrome is a distinct entity based on specific clinical features and unique pathobiology, its initial clinical manifestation can appear very similar to CIE, which would not prompt any further cardiac evaluation. The description of this novel syndrome emphasizes the critical, potentially life-saving importance of genetic diagnosis in patients with ichthyosis.\n\nMutations in ELOVL4 cause ichthyosis, intellectual disability, and spastic quadriplegia. In 2011, Aldamesh et al. identified a novel neuro-ichthyotic disease caused by autosomal recessive mutations in ELOVL4 and characterized by ichthyosis, mental retardation, seizures, and spastic quadriplegia71. These phenotypic manifestations are similar to those of Sjögren-Larrson syndrome, a syndromic ichthyosis caused by mutations in fatty aldehyde dehydrogenase (ALDH3A2)30, but the neurologic findings in this newly described disorder are more severe71.\n\nELOVL4 is a fatty acid elongase and plays a crucial role in the synthetic pathway of very-long-chain fatty acids (VLCFAs)72. VLCFAs have a wide range of functions, including cell signaling and maintenance of the epidermal barrier73–78. Heterozygous mutations in ELOVL4 had been previously reported to cause macular degeneration79, but Aldamesh et al. were first to report the involvement of homozygous ELOVL4 mutations in an ichthyosis syndrome, with both cutaneous and neurologic findings. In addition to identifying a novel syndromic ichthyosis, this finding highlights the importance of VLCFAs in both brain development and maintenance of epidermal barrier function. It also suggests the use of VLCFA replacement therapy as a possible therapeutic option for the treatment of patients with this disorder.",
"appendix": "Competing interests\n\n\n\nKeith Choate is a consultant for Aldeyra Therapeutics.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nOji V, Tadini G, Akiyama M, et al.: Revised nomenclature and classification of inherited ichthyoses: results of the First Ichthyosis Consensus Conference in Sorèze 2009. J Am Acad Dermatol. 2010; 63(4): 607–41. PubMed Abstract | Publisher Full Text\n\nFeingold KR: The outer frontier: the importance of lipid metabolism in the skin. J Lipid Res. 2009; 50(Suppl): S417–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMadison KC: Barrier function of the skin: \"la raison d'être\" of the epidermis. J Invest Dermatol. 2003; 121(2): 231–41. PubMed Abstract | Publisher Full Text\n\nNemes Z, Steinert PM: Bricks and mortar of the epidermal barrier. Exp Mol Med. 1999; 31(1): 5–19. PubMed Abstract | Publisher Full Text\n\nWells RS, Kerr CB: Clinical features of autosomal dominant and sex-linked ichthyosis in an English population. Br Med J. 1966; 1(5493): 947–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith FJ, Irvine AD, Terron-Kwiatkowski A, et al.: Loss-of-function mutations in the gene encoding filaggrin cause ichthyosis vulgaris. Nat Genet. 2006; 38(3): 337–42. PubMed Abstract | Publisher Full Text\n\nThyssen JP, Godoy-Gijon E, Elias PM: Ichthyosis vulgaris: the filaggrin mutation disease. Br J Dermatol. 2013; 168(6): 1155–66. PubMed Abstract | Publisher Full Text\n\nOsawa R, Akiyama M, Shimizu H: Filaggrin gene defects and the risk of developing allergic disorders. Allergol Int. 2011; 60(1): 1–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSandilands A, Terron-Kwiatkowski A, Hull PR, et al.: Comprehensive analysis of the gene encoding filaggrin uncovers prevalent and rare mutations in ichthyosis vulgaris and atopic eczema. Nat Genet. 2007; 39(5): 650–4. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nLykkesfeldt G, Nielsen MD, Lykkesfeldt AE: Placental steroid sulfatase deficiency: biochemical diagnosis and clinical review. Obstet Gynecol. 1984; 64(1): 49–54. PubMed Abstract\n\ndel Refugio Rivera Vega M, Murillo-Vilches MR, Toral-Lopez J, et al.: X-linked ichthyosis in a patient with a novel nonsense mutation in the STS gene. J Dermatol Sci. 2015; 80(2): 160–2. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRichard G, Bale SJ: Autosomal Recessive Congenital Ichthyosis. In: Pagon RA, Adam MP, Ardinger HH, Wallace SE, Amemiya A, Bean LJH, et al., editors. GeneReviews(R). Seattle (WA); 1993. PubMed Abstract\n\nMitsutake S, Suzuki C, Akiyama M, et al.: ABCA12 dysfunction causes a disorder in glucosylceramide accumulation during keratinocyte differentiation. J Dermatol Sci. 2010; 60(2): 128–9. PubMed Abstract | Publisher Full Text\n\nAkiyama M: ABCA12 mutations and autosomal recessive congenital ichthyosis: a review of genotype/phenotype correlations and of pathogenetic concepts. Hum Mutat. 2010; 31(10): 1090–6. PubMed Abstract | Publisher Full Text\n\nRajpopat S, Moss C, Mellerio J, et al.: Harlequin ichthyosis: a review of clinical and molecular findings in 45 cases. Arch Dermatol. 2011; 147(6): 681–6. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nTakeichi T, Akiyama M: Inherited ichthyosis: Non-syndromic forms. J Dermatol. 2016; 43(3): 242–51. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nFischer J: Autosomal recessive congenital ichthyosis. J Invest Dermatol. 2009; 129(6): 1319–21. PubMed Abstract | Publisher Full Text\n\nEckl KM, de Juanes S, Kurtenbach J, et al.: Molecular analysis of 250 patients with autosomal recessive congenital ichthyosis: evidence for mutation hotspots in ALOXE3 and allelic heterogeneity in ALOX12B. J Invest Dermatol. 2009; 129(6): 1421–8. PubMed Abstract | Publisher Full Text\n\nArin MJ, Oji V, Emmert S, et al.: Expanding the keratin mutation database: novel and recurrent mutations and genotype-phenotype correlations in 28 patients with epidermolytic ichthyosis. Br J Dermatol. 2011; 164(2): 442–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nCervantes T, Pham C, Browning JC: Superficial epidermolytic ichthyosis: a report of two families. Pediatr Dermatol. 2013; 30(4): 469–72. PubMed Abstract | Publisher Full Text\n\nRothnagel JA, Traupe H, Wojcik S, et al.: Mutations in the rod domain of keratin 2e in patients with ichthyosis bullosa of Siemens. Nat Genet. 1994; 7(4): 485–90. PubMed Abstract | Publisher Full Text\n\nYang JM, Yoneda K, Morita E, et al.: An alanine to proline mutation in the 1A rod domain of the keratin 10 chain in epidermolytic hyperkeratosis. J Invest Dermatol. 1997; 109(5): 692–4. PubMed Abstract | Publisher Full Text\n\nRichardson ES, Lee JB, Hyde PH, et al.: A novel mutation and large size polymorphism affecting the V2 domain of keratin 1 in an African-American family with severe, diffuse palmoplantar keratoderma of the ichthyosis hystrix Curth-Macklin type. J Invest Dermatol. 2006; 126(1): 79–84. PubMed Abstract | Publisher Full Text\n\nSprecher E, Ishida-Yamamoto A, Becker OM, et al.: Evidence for novel functions of the keratin tail emerging from a mutation causing ichthyosis hystrix. J Invest Dermatol. 2001; 116(4): 511–9. PubMed Abstract | Publisher Full Text\n\nFonseca DJ, Rojas RF, Vergara JI, et al.: A severe familial phenotype of Ichthyosis Curth-Macklin caused by a novel mutation in the KRT1 gene. Br J Dermatol. 2013; 168(2): 456–8. PubMed Abstract | Publisher Full Text\n\nKubo Y, Urano Y, Matsuda R, et al.: Ichthyosis hystrix, Curth-Macklin type: a new sporadic case with a novel mutation of keratin 1. Arch Dermatol. 2011; 147(8): 999–1001. PubMed Abstract | Publisher Full Text\n\nMirza H, Kumar A, Craiglow BG, et al.: Mutations Affecting Keratin 10 Surface-Exposed Residues Highlight the Structural Basis of Phenotypic Variation in Epidermolytic Ichthyosis. J Invest Dermatol. 2015; 135(12): 3041–50. PubMed Abstract | Publisher Full Text\n\nLai-Cheong JE, Elias PM, Paller AS: Pathogenesis-based therapies in ichthyoses. Dermatol Ther. 2013; 26(1): 46–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCanueto J, Girós M, Ciria S, et al.: Clinical, molecular and biochemical characterization of nine Spanish families with Conradi-Hünermann-Happle syndrome: new insights into X-linked dominant chondrodysplasia punctata with a comprehensive review of the literature. Br J Dermatol. 2012; 166(4): 830–8. PubMed Abstract | Publisher Full Text\n\nGaboon NE, Jelani M, Almramhi MM, et al.: Case of Sjögren-Larsson syndrome with a large deletion in the ALDH3A2 gene confirmed by single nucleotide polymorphism array analysis. J Dermatol. 2015; 42(7): 706–9. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nGrall A, Guaguère E, Planchais S, et al.: PNPLA1 mutations cause autosomal recessive congenital ichthyosis in golden retriever dogs and humans. Nat Genet. 2012; 44(2): 140–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBaulande S, Langlois C: [Proteins sharing PNPLA domain, a new family of enzymes regulating lipid metabolism]. Med Sci (Paris). 2010; 26(2): 177–84. PubMed Abstract | Publisher Full Text\n\nKienesberger PC, Oberer M, Lass A, et al.: Mammalian patatin domain containing proteins: a family with diverse lipolytic activities involved in multiple biological functions. J Lipid Res. 2009; 50(Suppl): S63–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFischer J, Lefèvre C, Morava E, et al.: The gene encoding adipose triglyceride lipase (PNPLA2) is mutated in neutral lipid storage disease with myopathy. Nat Genet. 2007; 39(1): 28–30. PubMed Abstract | Publisher Full Text\n\nMubaidin A, Roberts E, Hampshire D, et al.: Karak syndrome: a novel degenerative disorder of the basal ganglia and cerebellum. J Med Genet. 2003; 40(7): 543–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRainier S, Bui M, Mark E, et al.: Neuropathy target esterase gene mutations cause motor neuron disease. Am J Hum Genet. 2008; 82(3): 780–5. PubMed Abstract | Publisher Full Text\n\nRomeo S, Kozlitina J, Xing C, et al.: Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat Genet. 2008; 40(12): 1461–5. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBoyden LM, Craiglow BG, Zhou J, et al.: Dominant De Novo Mutations in GJA1 Cause Erythrokeratodermia Variabilis et Progressiva, without Features of Oculodentodigital Dysplasia. J Invest Dermatol. 2015; 135(6): 1540–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaird DW: Syndromic and non-syndromic disease-linked Cx43 mutations. FEBS Lett. 2014; 588(8): 1339–48. PubMed Abstract | Publisher Full Text\n\nMartin PE, Easton JA, Hodgins MB, et al.: Connexins: sensors of epidermal integrity that are therapeutic targets. FEBS Lett. 2014; 588(8): 1304–14. PubMed Abstract | Publisher Full Text\n\nPfenniger A, Wohlwend A, Kwak BR: Mutations in connexin genes and disease. Eur J Clin Invest. 2011; 41(1): 103–16. PubMed Abstract | Publisher Full Text\n\nMacari F, Landau M, Cousin P, et al.: Mutation in the gene for connexin 30.3 in a family with erythrokeratodermia variabilis. Am J Hum Genet. 2000; 67(5): 1296–301. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichard G, Smith LE, Baily RA, et al.: Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis. Nat Genet. 1998; 20(4): 366–9. PubMed Abstract | Publisher Full Text\n\nPaznekas WA, Boyadjiev SA, Shapiro RE, et al.: Connexin 43 (GJA1) mutations cause the pleiotropic phenotype of oculodentodigital dysplasia. Am J Hum Genet. 2003; 72(2): 408–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShao Q, Liu Q, Lorentz R, et al.: Structure and functional studies of N-terminal Cx43 mutants linked to oculodentodigital dysplasia. Mol Biol Cell. 2012; 23(17): 3312–21. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nAlbert MR, Mackool BT: Pityriasis rubra pilaris. Int J Dermatol. 1999; 38(1): 1–11. PubMed Abstract | Publisher Full Text\n\nBraun-Falco O, Ryckmanns F, Schmoeckel C, et al.: Pityriasis rubra pilaris: a clinico-pathological and therapeutic study with special reference to histochemistry, autoradiography, and electron microscopy. Arch Dermatol Res. 1983; 275(5): 287–95. PubMed Abstract | Publisher Full Text\n\nMagro CM, Crowson AN: The clinical and histomorphological features of pityriasis rubra pilaris. A comparative analysis with psoriasis. J Cutan Pathol. 1997; 24(7): 416–24. PubMed Abstract | Publisher Full Text\n\nSoeprono FF: Histologic criteria for the diagnosis of pityriasis rubra pilaris. Am J Dermatopathol. 1986; 8(4): 277–83. PubMed Abstract | Publisher Full Text\n\nAuffret N, Quint L, Domart P, et al.: Pityriasis rubra pilaris in a patient with human immunodeficiency virus infection. J Am Acad Dermatol. 1992; 27(2 Pt 1): 260–1. PubMed Abstract | Publisher Full Text\n\nKlein A, Landthaler M, Karrer S: Pityriasis rubra pilaris: a review of diagnosis and treatment. Am J Clin Dermatol. 2010; 11(3): 157–70. PubMed Abstract | Publisher Full Text\n\nMenni S, Brancaleone W, Grimalt R: Pityriasis rubra pilaris in a child seropositive for the human immunodeficiency virus. J Am Acad Dermatol. 1992; 27(6 Pt 1): 1009. PubMed Abstract | Publisher Full Text\n\nGriffiths WA: Pityriasis rubra pilaris. Clin Exp Dermatol. 1980; 5(1): 105–12. PubMed Abstract | Publisher Full Text\n\nVasher M, Smithberger E, Lien MH, et al.: Familial pityriasis rubra pilaris: report of a family and therapeutic response to etanercept. J Drugs Dermatol. 2010; 9(7): 844–50. PubMed Abstract\n\nFuchs-Telem D, Sarig O, van Steensel MA, et al.: Familial pityriasis rubra pilaris is caused by mutations in CARD14. Am J Hum Genet. 2012; 91(1): 163–70. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBertin J, Wang L, Guo Y, et al.: CARD11 and CARD14 are novel caspase recruitment domain (CARD)/membrane-associated guanylate kinase (MAGUK) family members that interact with BCL10 and activate NF-kappa B. J Biol Chem. 2001; 276(15): 11877–82. PubMed Abstract | Publisher Full Text\n\nScudiero I, Zotti T, Ferravante A, et al.: Alternative splicing of CARMA2/CARD14 transcripts generates protein variants with differential effect on NF-κB activation and endoplasmic reticulum stress-induced cell death. J Cell Physiol. 2011; 226(12): 3121–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJordan CT, Cao L, Roberson ED, et al.: PSORS2 is due to mutations in CARD14. Am J Hum Genet. 2012; 90(5): 784–95. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nArita K, Jacyk WK, Wessagowit V, et al.: The South African \"bathing suit ichthyosis\" is a form of lamellar ichthyosis caused by a homozygous missense mutation, p.R315L, in transglutaminase 1. J Invest Dermatol. 2007; 127(2): 490–3. PubMed Abstract | Publisher Full Text\n\nJacyk WK: Bathing-suit ichthyosis. A peculiar phenotype of lamellar ichthyosis in South African blacks. Eur J Dermatol. 2005; 15(6): 433–6. PubMed Abstract\n\nOji V, Hautier JM, Ahvazi B, et al.: Bathing suit ichthyosis is caused by transglutaminase-1 deficiency: evidence for a temperature-sensitive phenotype. Hum Mol Genet. 2006; 15(21): 3083–97. PubMed Abstract | Publisher Full Text\n\nBenmously-Mlika R, Zaouak A, Mrad R, et al.: Bathing suit ichthyosis caused by a TGM1 mutation in a Tunisian child. Int J Dermatol. 2014; 53(12): 1478–80. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBourrat E, Blanchet-Bardon C, Derbois C, et al.: Specific TGM1 mutation profiles in bathing suit and self-improving collodion ichthyoses: phenotypic and genotypic data from 9 patients with dynamic phenotypes of autosomal recessive congenital ichthyosis. Arch Dermatol. 2012; 148(10): 1191–5. PubMed Abstract | Publisher Full Text\n\nYamamoto M, Sakaguchi Y, Itoh M, et al.: Bathing suit ichthyosis with summer exacerbation: a temperature-sensitive case. Br J Dermatol. 2012; 166(3): 672–4. PubMed Abstract | Publisher Full Text\n\nBoyden LM, Kam CY, Hernandez-Martin A, et al.: Dominant de novo DSP mutations cause erythrokeratodermia-cardiomyopathy syndrome. Hum Mol Genet. 2016; 25(2): 348–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArmstrong DK, McKenna KE, Purkis PE, et al.: Haploinsufficiency of desmoplakin causes a striate subtype of palmoplantar keratoderma. Hum Mol Genet. 1999; 8(1): 143–8. PubMed Abstract | Publisher Full Text\n\nPigors M, Schwieger-Briel A, Cosgarea R, et al.: Desmoplakin mutations with palmoplantar keratoderma, woolly hair and cardiomyopathy. Acta Derm Venereol. 2015; 95(3): 337–40. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRampazzo A, Nava A, Malacrida S, et al.: Mutation in human desmoplakin domain binding to plakoglobin causes a dominant form of arrhythmogenic right ventricular cardiomyopathy. Am J Hum Genet. 2002; 71(5): 1200–6. PubMed Abstract | Publisher Full Text\n\nvan der Zwaag PA, Jongbloed JD, van den BMP, et al.: A genetic variants database for arrhythmogenic right ventricular dysplasia/cardiomyopathy. Hum Mutat. 2009; 30(9): 1278–83. PubMed Abstract | Publisher Full Text\n\nWhittock NV, Ashton GH, Dopping-Hepenstal PJ, et al.: Striate palmoplantar keratoderma resulting from desmoplakin haploinsufficiency. J Invest Dermatol. 1999; 113(6): 940–6. PubMed Abstract | Publisher Full Text\n\nAldahmesh MA, Mohamed JY, Alkuraya HS, et al.: Recessive mutations in ELOVL4 cause ichthyosis, intellectual disability, and spastic quadriplegia. Am J Hum Genet. 2011; 89(6): 745–50. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nJakobsson A, Westerberg R, Jacobsson A: Fatty acid elongases in mammals: their regulation and roles in metabolism. Prog Lipid Res. 2006; 45(3): 237–49. PubMed Abstract | Publisher Full Text\n\nLeonard AE, Kelder B, Bobik EG, et al.: Identification and expression of mammalian long-chain PUFA elongation enzymes. Lipids. 2002; 37(8): 733–40. PubMed Abstract | Publisher Full Text\n\nMcMahon A, Butovich IA, Mata NL, et al.: Retinal pathology and skin barrier defect in mice carrying a Stargardt disease-3 mutation in elongase of very long chain fatty acids-4. Mol Vis. 2007; 13: 258–72. PubMed Abstract | Free Full Text\n\nPoulos A, Beckman K, Johnson DW, et al.: Very long-chain fatty acids in peroxisomal disease. Adv Exp Med Biol. 1992; 318: 331–40. PubMed Abstract\n\nSchneiter R, Brugger B, Amann CM, et al.: Identification and biophysical characterization of a very-long-chain-fatty-acid-substituted phosphatidylinositol in yeast subcellular membranes. Biochem J. 2004; 381(Pt 3): 941–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToulmay A, Schneiter R: Lipid-dependent surface transport of the proton pumping ATPase: a model to study plasma membrane biogenesis in yeast. Biochimie. 2007; 89(2): 249–54. PubMed Abstract | Publisher Full Text\n\nVasireddy V, Uchida Y, Salem N, et al.: Loss of functional ELOVL4 depletes very long-chain fatty acids (or =C28) and the unique omega-O-acylceramides in skin leading to neonatal death. Hum Mol Genet. 2007; 16(5): 471–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang K, Kniazeva M, Han M, et al.: A 5-bp deletion in ELOVL4 is associated with two related forms of autosomal dominant macular dystrophy. Nat Genet. 2001; 27(1): 89–93. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "14586",
"date": "24 Jun 2016",
"name": "Masashi Akiyama",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14588",
"date": "24 Jun 2016",
"name": "Anna Bruckner",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1497
|
https://f1000research.com/articles/5-1493/v1
|
24 Jun 16
|
{
"type": "Review",
"title": "Tourette syndrome research highlights 2015",
"authors": [
"Cheryl A. Richards",
"Kevin J. Black",
"Kevin J. Black"
],
"abstract": "We present selected highlights from research that appeared during 2015 on Tourette syndrome and other tic disorders. Topics include phenomenology, comorbidities, developmental course, genetics, animal models, neuroimaging, electrophysiology, pharmacology, and treatment. We briefly summarize articles whose results we believe may lead to new treatments, additional research or modifications in current models of TS.",
"keywords": [
"Tourette syndrome",
"tic disorders",
"genetics",
"animal models",
"neuroimaging",
"premonitory",
"therapy",
"review"
],
"content": "Introduction\n\nThis article is the second in the TS Research Highlights series1. These articles are meant to disseminate recent scientific progress on Gilles de la Tourette Syndrome (TS). During each year, the article will be a work in progress, maintained as a web page on the Authorea online authoring platform (the working draft for 2016 appears here). After the calendar year ends, the article is finalized and submitted as the annual update for the Tics channel on F1000Research2.\n\n\nMethods\n\nWe searched PubMed on 22 Jan 2016 using the search strategy: (“Tic Disorders”[MeSH] OR Tourette NOT Tourette[AU]) AND 2015[PDAT]. This search returned 202 citations and includes articles appearing online in 2015 but not officially published by year end (from journals that still focus on the paper user interface). We also reviewed F1000Prime recommendations and presentations of interest at selected medical conferences. Articles were chosen based on a purely subjective assessment of interest, guided by our judgment of possible future impact on the field.\n\n\nResults\n\nTic suppression. Since tic suppression is part of the treatment protocol for the Comprehensive Behavioral Intervention for Tics (CBIT) and for Exposure and Response Prevention, there has been increased interest in investigating the characteristics of tic suppression and the factors that affect it. A study of 26 TS adolescents compared free ticcing with a tic suppression condition3. During the free ticcing condition, tic distribution across body locations was consistent with the view that most tics occur at the level of the shoulders and above: eye tics were the most frequent, followed by facial/cervical tics, and those involving the arms and legs. Tics involving the trunk were the least common. During the tic suppression condition, eye tics increased in 10 subjects, as did hand tics in 3 subjects. Tic suppression was most successful for tics in body locations generally associated with fewer tics, such as the legs and trunk. The authors suggest that tic suppression involves specific, rather than global, inhibition since some types of tics were easier to suppress than others. Historically, other categories have been used to classify tics, such as simple vs. complex tics and motor vs. phonic tics. The results of this study suggest that future research may benefit from including body location in tic analyses.\n\nBy definition, children with Tourette syndrome (TS) have had tics for over a year. They can often suppress their tics briefly and they do so more effectively when rewarded for successful suppression. It has not been known whether the ability to suppress tics develops only with practice over the years of having tics or whether the ability to suppress tics is present when tics initially occur. Greene and colleagues addressed this question in children whose tics had developed within the past few months4. When children received tokens with monetary value for tic-free intervals, they had significantly more of these intervals compared to a baseline, unrewarded condition. This result suggests the possibility that behavior therapy for tics may work, at least for some children, even before TS can be diagnosed.\n\nAnother study examined the association between tic suppression and quality of life. Although most tic patients frequently try to suppress tics, they find suppressing them uncomfortable and distracting. However, those patients who are more satisfied with their ability to suppress their tics also report a higher quality of life5.\n\nSensory phenomena. Premonitory urges are common in TS, are perceived as triggers for tics, and are as bothersome as the tics themselves to many people with TS. Premonitory urges have a sensory component and many TS patients also report sensory sensitivities. Researchers have been clarifying the nature of premonitory urges and attempting to determine the underlying causes of sensory sensitivities.\n\nScores on the Premonitory Urge for Tics Scale (PUTS) and the University of São Paulo Sensory Phenomena Scale (USP-SPS) were significantly correlated with total tic severity, tic complexity and vocal tic scores in youth and adults with TS6. The PUTS and USP-SPS scores were also correlated with scores on the Dimensional Yale-Brown Obsessive-Compulsive Scale. This study provides additional evidence that the association between premonitory urges and tics is complex and may be influenced by obsessive-compulsive tendencies.\n\nAnother study examined the association between premonitory urges and interoceptive awareness7. Interoceptive awareness was measured by how well subjects were able to mentally track their heartbeats, without being able to take their own pulse, during a specific period of time. Interoceptive awareness, tic severity, and obsessive-compulsive symptom severity were used in a multiple regression to predict PUTS scores. Greater interoceptive awareness and tic severity were significantly associated with higher PUTS scores while OCD symptom severity was not. The authors suggest that high interoceptive awareness may lead people to set a low threshold for perception of their own internal physiological sensations and therefore cause them to interpret these sensations as an urge to tic. Interestingly, TS subjects had lower interoceptive awareness than controls overall, suggesting that this downregulation of interoception might reflect a compensatory process.\n\nSymptoms and comorbidity. Recent research has again demonstrated the wide prevalence of TS-associated comorbidities and is a reminder of the need to perform studies with large enough sample sizes to examine the effects of comorbidities on the dependent variables of interest.\n\nA retrospective review of 1,000,000 people in the Taiwan National Health Insurance Research Database examined the association between epilepsy and TS8. 1,062 children and adolescents with TS were matched on age and sex with a control group of 3,186 subjects. The TS group had an 18-fold increased risk of epilepsy compared to the control group; the risk was still elevated 16-fold after adjusting for comorbidities (i.e., bipolar disorder, depression, learning difficulties, autism, anxiety disorders, sleep disorders). The authors acknowledge the possibility that some tics may have been mistaken for seizures, given the nature of the data set, but these data do suggest an important hypothesis for future confirmation.\n\nIn a large study of psychiatric comorbidity in TS, approximately 800 families were recruited primarily from TS specialty clinics in four different countries over a 16-year period9. A total of 1374 participants with TS and 1142 family members unaffected by TS were included. 86% of the TS participants had at least one psychiatric comorbidity and 72% had either OCD or ADHD. Mood, anxiety and disruptive behavior disorders each occurred in approximately 30% of the TS participants. The genetic correlations between TS and mood were accounted for by ADHD and OCD, while ADHD alone accounted for the genetic correlations of TS with anxiety and disruptive behavior disorders.\n\nA study of 400 patients seen at a TS specialty clinic found that 39% had coprolalia and 20% had copropraxia10. When the 222 patients with full comorbidity data were examined, only 13.5% had “pure” TS (i.e., without comorbidities). None of the “pure” TS group had coprolalia and none had a family history of obsessive-compulsive disorder.\n\nEmotional regulation difficulties were described in three studies, reminding us that for many TS patients, tics are not their most problematic symptoms. Greater irritability was seen in TS adults with more severe tics and those with comorbid ADHD11. Eddy et al. found that both male and female TS subjects, compared to controls, reported more distress during emotionally intense situations and rated their abilities to take other people’s perspectives lower12. An experienced clinician who has done research on “rage attacks” in TS has provided a clinically useful summary of current knowledge regarding aggressive symptoms in TS, OCD, ADHD and mood disorders, and described treatment options13. Given that emotional regulation difficulties are frequently associated with greater tic severity, improving emotional modulation skills may be an appropriate target of psychological interventions.\n\nMore research is also being conducted on personality differences associated with TS. A small study of 17 male adolescents14 found that the only significant difference between the TS subjects and 51 age- and gender-matched controls on the Minnesota Multiphasic Personality Inventory-Adolescent version was that the TS subjects scored higher on the Obsessiveness Content Scale. In contrast, a study of 50 TS adults in Germany used a variety of instruments to measure psychological symptoms and personality traits15. Comorbidities were common (41% OCD, 28% depression, 26% ADHD). Patients with OCD had more severe tics and there was a trend for those with ADHD to have more severe tics. Only 29% of the patients had no pathological personality traits, as measured by the Inventory of Clinical Personality Accentuations. The demand-anxious trait was the most common personality trait seen in patients and was present in 39%, while histrionic personality traits were not found in any patients. Personality traits in patients with “pure” TS were comparable to those of the control group. Interestingly, ADHD did not contribute to increased probability of pathological personality traits. Although quality of life was affected by both personality traits and comorbidities, personality traits had a larger impact on quality of life.\n\nCourse. More research is being conducted on the developmental progression of tics and other symptoms in TS. This work may provide clues that help clarify what factors contribute to the appearance and disappearance of transient tics and what factors explain the disappearance or amelioration of tics as children with TS enter adulthood.\n\nOne study examined home videos recorded in the first 6 months of life from 34 children who were exhibiting autistic behaviors in their second year of life16. Families reported that development during the first year of life had been normal. Videos of 18 boys were examined in detail. The primary focus of the study was on autism, though 11 of the 18 subjects were later diagnosed with TS. The nearly ubiquitous availability of home baby videos in some cultures suggests that a similar pseudo-prospective study design could be used to identify behavioral features predicting later development of TS.\n\nThe clinical characteristics of children who developed TS before the age of 4 were compared with those whose tics developed at age 6 or older17. The younger group had a higher rate of speech dysfluencies (e.g., stuttering) and oppositional defiant disorder. There was no difference between the two groups in prevalence of ADHD or obsessive-compulsive symptoms. Interestingly, the children in the early-onset group were more likely to have mothers with tics. The authors attributed this to mothers with tics being more likely to recognize tics in their children. The authors also suggest that prenatal or perinatal maternal environmental factors may contribute to the development of tics. An alternative explanation may relate to the fact that TS is much less common in girls than in boys. Consequently, tics in a woman may represent a higher genetic load, resulting in a more severe form of tics and an earlier age of onset in her children.\n\nResearchers re-evaluated 75 patients previously seen at a university-based TS clinic after a mean follow-up of 9 years18. Reported TS impairment was more likely to decrease over time in males and increase in females. In adulthood, women were more likely than men to have an expansion of the number of body regions exhibiting tics, primarily in the upper extremities. This result suggests that sex continues to influence TS symptoms beyond adolescence.\n\nGenetics. The most important genetics news of the year may have been the presentation by Huang and colleagues at the Tourette World Congress in London in June, 2015, reporting on a large collaborative study of copy number variants (CNVs) in approximately 2,500 TS cases and 3,500 controls19. They first tested some CNVs previously reported in various neurological and psychiatric illnesses; the most significant confirmation in this sample was of exonic deletions in NRXN1, a gene previously implicated in autism and TS20. They also searched for new, large CNVs and identified a novel TS locus, CNTN6, that was significant at a whole-genome level by permutation testing. This gene and 4 of the 32 next most likely candidates this analysis identified are neural adhesion molecules20. CNTN6 is a reasonable candidate for etiologically contributing to TS: its expression in the brain varies during development21, knockout mice show motor deficits22, and an independent study found a variety of neurological and psychiatric symptoms (including 2 with OCD) in 14 patients identified by CNTN6 CNVs23. On the other hand, fewer than 1% of TS cases in this sample had CNVs in CNTN6, so its overall importance in TS remains to be more fully characterized.\n\nA large collaborative group24 studied AADAC, the gene encoding arylacetamide deacetylase, in which microdeletions had been identified in a previous, smaller copy number variation study. The authors provide evidence from several sources supporting the connection of AADAC and TS, including a significant overall association, new patients with AADAC deletions, and evidence that AADAC is expressed in the brain, though most strongly in cerebellum, hippocampus and olfactory bulb, rather than the basal ganglia.\n\nA collaborative genetic study25 demonstrated an association of TS with 33 genes related to glycolysis or glutamine metabolism. None of the individual genes would have survived correction for multiple testing and the results were consistent with a combined effect of many genetic variants of small effect. These results suggest a new direction for future genetic, electrophysiological, imaging and pharmacological studies.\n\nYu et al. reported a genome-wide association study (GWAS) from 1,310 people with OCD, 834 with Tourette syndrome, 579 with both OCD and a chronic tic disorder, and over 5,500 controls matched for ancestry26. A significant polygenic component was identified for OCD without tics, but not for the combined patient group or other subgroups. Overall, this study is consistent with previous work but it provided disappointingly few novel results.\n\nAn international study examined tic symptoms in the United States and the Netherlands27. Three factors (complex vocal tics and obscene behavior, body tics, and head/neck tics) accounted for 49% of the variance in tic-related symptoms. There was no evidence of heritability for the second factor, but the h2r was approximately 0.2 for the first and third factors when age and sex were included as covariates. Heritability for these narrower tic phenotypes is considerably lower than the heritability estimates (up to 0.65) when comorbid conditions such as OCD and ADHD are included. These authors conclude that broader tic phenotypes, rather than narrower pure tic phenotypes, may be more successful at identifying the genetic mechanisms underlying TS.\n\nEnvironmental risk factors. Researchers used data from the Avon Longitudinal Study of Parents and Children to identify maternal factors that increase the risk of tic disorders in offspring28. The Avon Longitudinal Study is an ongoing, prospective, pre-birth cohort study of all children born in Avon, United Kingdom, between April 1, 1991, and December 31, 1992. Maternal questionnaires were administered throughout pregnancy and mothers also completed questionnaires about themselves and their children’s development every 6 months from the child’s birth to the age of 7 and then yearly thereafter. In the final multivariate model, chronic maternal anxiety, evident both before and after the child’s birth, was associated with TS or chronic tic disorder in children. This association may reflect shared genetic susceptibility or prenatal exposure.\n\nPathological studies. An important study follows up on the autopsy results from the Vaccarino lab by comparing RNA transcripts from the basal ganglia of 9 TS and 9 matched control subjects29. The most strongly associated set of downregulated transcripts involved striatal interneurons, consistent with the autopsy studies. The leading set of upregulated transcripts involved immune-related genes even though none of the TS subjects met proposed diagnostic criteria for PANDAS or PANS. The results obtained in the present study using brain tissue did not overlap with those of previous studies using blood samples. The authors interpret their results as implicating disrupted basal ganglia interneuron signaling in the pathophysiology of severe TS.\n\nAnimal models. Rodent and monkey tic models have been developed in order to study tic generation mechanisms more directly and a number of studies using rodent models were published in 2015. Removing about half of the cholinergic interneurons in the dorsolateral striatum produced increased fragmented grooming behavior in response to repeated unpredictable acoustic startle stimuli in mice and also increased repetitive sniffing in response to D-amphetamine challenge30. Ablation in the dorsomedial striatum did not produce similar deficits. None of the experimental conditions produced a change in prepulse inhibition. These results provide partial support for the autopsy data linking some characteristic TS symptoms to cholinergic interneuron deficits in the dorsolateral striatum.\n\nA rat model was used to determine to what extent cortical and striatal input affected the temporal and spatial properties of motor tics31. Focal blockade of GABA-A receptors with bicuculline injections into the anterior striatal motor region produced focal tic-like movements of the forelimbs. Medium spiny neurons (MSNs) and fast spiking interneurons (FSIs) exhibited increased activity during these movements, and all of the MSNs were only active during them. All of the FSIs were active during tics, but a minority followed this increase in activity by a decrease. Four different patterns were seen in globus pallidus (GP) neurons. About half of the GP neurons demonstrated increased activity during the tic-like movements, while the rest showed only inhibition or a combination of inhibition and excitation. The effects of cortical input were studied using short bursts of high-frequency electrical pulses applied at random intervals to the region of primary motor cortex representing the forelimb. Stimulation was provided before and after the bicuculline injections. The results suggested that the precise timing of tic occurrence was related to both incoming excitatory cortical input and the delay since the previous tic. These results support the fundamental involvement of the corticostriatal network with tic occurrence.\n\nThe role of GABA in tic generation was also studied in adult mice by injecting the GABA-A antagonist picrotoxin into areas throughout the cortex and striatum32. Infusions into the central and dorsolateral striatum produced tic-like movements of the front paw, hind paw or head. Infusions into the dorsomedial striatum did not have a significant behavioral effect. Infusion into the ventral striatum produced increased locomotor activation in addition to sterotypical sniffing and wall licking without other tic-like movements. Infusions into the sensorimotor cortex produced tic-like movements in addition to increased behavioral activation involving cage exploration, sniffing, and occasional licking. When an NMDA receptor antagonist was infused into the dorsolateral striatum prior to infusing picrotoxin into the same location, tic frequency decreased significantly, thus demonstrating a role of glutamatergic activity in tic generation. Infusion of a GABA-A agonist into the sensorimotor cortex prior to picrotoxin infusion in the dorsolateral striatum also resulted in significant tic suppression. EEG recordings ruled out seizure activity. The authors summarize these results as providing evidence that these tic-like movements require corticostriatal interactions, with a key role for glutatmateric afferents, rather than autonomous striatal activity.\n\nIn a genetically engineered mouse model of TS+OCD (“Ticcy” D1CT-7 transgenic mice)33, a small population of dopamine D1 receptor-expressing (D1+) somatosensory cortical and limbic neurons is chronically potentiated, resulting in cortical and amygdalar glutamatergic excitation of striatothalamic, striatopallidal and nigrostriatal subcircuits. Tics were decreased by the use of drugs that acted at different points in this “hyperglutamergic cortico/amygdalo-striato-thalamo-cortical [CSTC] circuit”. Excitatory forebrain serotonin and norepinephrine activity was blocked by ritanserin (a serotonin 2a/2c antagonist) and prazosin (an α1 adrenergic antagonist) respectively. In contrast, downstream striatothalamic neurons’ glutamate-triggered GABA output and downstream nigrostriatal neurons’ glutamate-triggered co-modulatory dopamine output were blocked by moxonidine (an agmatine/imidazoline-1 agonist) and bromocriptine (a D2 dopamine agonist) respectively. All four of these drugs decreased tic frequency and were considered to be “circuit-breakers” for the hyperglutamatergic CSTC circuit, thus supporting an important role of glutamate in generating the abnormal tic-like movements seen in these mice.\n\nNeuroimaging studies. The challenges of using neuroimaging techniques to study pediatric and clinical subjects are described in detail along with suggestions concerning various strategies that can be used to collect higher quality data34. The profound effects of even very small head movements on structural MRI analyses were identified in a well-designed study35. T1-weighted MRI of brain was acquired in 12 healthy adults while they were still or engaged in specific types of movement including nodding, head shaking or a movement each subject invented and then repeated during the scan run. Even during scans when subjects attempted to remain still, there was an average of 3 mm/min of accumulated motion measured as root mean square displacement per minute. During the motion conditions, substantial impact was found on gray matter volume and thickness estimates. Apparent volume loss averaged approximately 0.7% mm/min of subject motion. The greatest reductions in gray matter occurred in pre- and post-central cortex and in the temporal lobes. Motion-associated increases in thickness were seen in some frontal regions and deep sulci such as the medial orbital frontal region. Significant effects due to motion were still present even after excluding scans that failed a rigorous quality control procedure. Recommendations included reducing head motion during scans as much as possible, controlling for motion in statistical analyses, and using correlational analyses to determine the associations between head motion and the predictors of interest. Tisdall and colleagues described a method to limit the effects of movement artifacts by using a motion tracking system to provide prospective motion correction during scanning36.\n\nA whole-brain analysis of cortical gray matter found reduced gray matter (GM) thickness in the insula and sensorimotor cortex for 29 TS children and young adults compared to a matched control group37. GM thickness in these areas correlated negatively with Premonitory Urge for Tics Scale scores.\n\nResting-state functional magnetic resonance imaging identified greater functional connectivity between the right dorsal anterior insula (dAI) and the bilateral supplementary motor area (SMA) in TS adults compared to controls38. Post-hoc analyses found significant correlations between PUTS scores and connectivity between right dAI and right SMA2 and between right dAI and left SMA1. These regions may be involved in the increased awareness of body sensations that tend to be associated with premonitory urges. The authors paid attention to head movement and removed high-movement frames. However, recent result suggest that the motion threshold of 0.4mm used in this analysis, and the choice not to regress global signal, may not adequately remove artifactual correlations between brain regions due to residual small head movements during the scan39.\n\nA review summarized TS task-based fMRI studies in TS including studies of tic suppression, voluntary motor execution, voluntary motor inhibition, and tic severity40. Free ticcing conditions (four studies) most commonly activated the left cerebellum, right cingulum, left middle frontal gyrus, the Rolandic operculum, right pallidum, right SMA and thalamus. In motor response inhibition studies, on No-Go trials TS subjects exhibited greater activation in the bilateral prefrontal cortex, thalamus and caudate. In contrast, on voluntary motor execution tasks greater activation in TS subjects was seen in the left prefrontal cortex, right cingulum, and the anterior SMA. Tic severity ratings were correlated with greater activation of the right dorsal premotor cortex and the SMA. Anterior cingulate cortex and SMA were involved across task types. The thalamus was involved in all types of studies except for self-produced movements. The authors also briefly summarize the many issues related to neuroimaging tasks, such as associated comorbidities, medication effects, the need for longitudinal studies, and the confounding effect of tics during scanning. Additional neuroimaging studies of note are noted in Table 1.\n\nElectrophysiology. Local field potentials associated with spontaneous tics were studied in 3 patients during DBS surgery46. In all 3 patients repetitive thalamo-cortical coherent activity was present from 800 to 1500 msec prior to tic-associated muscle contractions. The frequency range affected varied among the patients and there were also ongoing intermittent intra-thalamic coherences that were not synchronized to the tics. The authors speculated that specific DBS targets may not matter as much as whether the target is part of the striato-pallido-thalamo-cortical network. However, since these patients were older and had very severe and complex tics, the authors acknowledge that it is not yet clear to what extent these results generalize to the TS population as a whole. Additional studies are noted in Table 1.\n\nPharmacological studies. GABA involvement was studied in 23 TS children, aged 8–12, and 67 controls using a battery of vibrotactile tasks with a subset of the children (19 with TS, 25 controls) also undergoing GABA-edited magnetic resonance spectroscopy (MRS)47. Lower GABA concentration in the right sensorimotor cortex correlated with greater motor tic severity (r = –0.55). There were no significant differences between groups on reaction time and baseline amplitude discrimination threshold. However, TS children showed impaired tactile adaptation. The authors suggest that MRS GABA and tactile measures might useful as biomarkers of treatment response.\n\nPositron emission tomography (PET) was used to investigate striatal D2/D3 dopamine receptor availability in TS subjects and controls using a D2/D3 receptor antagonist ([11C]raclopride) and an agonist with preferential binding to D3Rs ([11C](+)PHNO)48. No differences were found in striatal regions when the TS subjects were compared with the controls, and there were no significant correlations between receptor availability and tic severity. The authors concluded that their results challenge the widely held view that striatal dopamine receptors have a fundamental role in TS pathophysiology, though they acknowledged that endogenous dopamine levels may have influenced the results since these radiotracers are displaceable by physiological synaptic concentrations of dopamine.\n\nTwo studies examined dopamine D1 and D2-like receptors in healthy adults, measuring D1R and D2R availability during stop-signal and continuous performance tasks49. Stop-signal reaction time was negatively correlated with both D1R and D2R receptor availability in the the associative and sensory motor regions of the striatum. In contrast, neither D1R nor D2R receptor availability was associated with performance on the continuous performance task, suggesting that stop-signal and continuous performance tasks are associated with different neurochemical mechanisms related to motor response inhibition. In a study of healthy adults, learning from positive outcomes was positively correlated with D1R binding in the putamen and caudate while there was an inverted U-shaped relationship (r2 = 0.19) between learning from negative outcomes and D2R binding in the putamen50. A dietary manipulation that reduced dopamine precursor levels significantly improved learning from negative outcomes. These results were interpreted as providing evidence that dopamine acts as a reward prediction error signal rather than as a saliency signal.\n\nA detailed review focuses on histaminergic modulation of striatal function and its possible role in TS51. The authors suggest that during wakefulness and increased attention, histaminergic neurons will be more active with the result that the striatum will be more responsive to thalamostriatal input and feed-forward inhibition will dominate. Several lines of evidence related to the role of histamine in TS were discussed. A family linkage study identified a rare mutation in the gene encoding histidine decarboxylase. HDC transgenic mice exhibit decreased pre-pulse inhibition of startle responses and an increase in a variety of amphetamine-induced stereotypies that were prevented by pretreatment with histamine infusion or use of haloperidol. Reduced histamine production was suggested to produce dopaminergic disregulation of the basal ganglia and symptoms similar to those seen in TS.\n\nClinical and neuropsychological studies. In an intriguing report from a group studying social cognition, TS subjects exhibited intact mentalizing when observing animated triangles demonstrating simple and complex interactions52. However, unlike controls, TS subjects also tended to attribute human-like intentions when two triangles were moving randomly. This tendency was not explained by clinical symptoms or by other constructs such as executive function or alexithymia.\n\nTwo studies examined the effects of comorbidities on social and cognitive skills. Social responsiveness and cognitive flexibility were examined in TS children and adolescents53. TS subjects were rated as having poorer social motivation and skills, using the Social Responsiveness Scale, compared to age-matched controls. TS subjects also took significantly longer to complete the Trail Making Tests which measure cognitive flexibility and visual motor integration. However, of the 31 TS subjects, 11 had OCD, 18 had ADHD and 8 had an anxiety disorder. Once these comorbidities were taken into account, group differences on the Trail Making Tests and the Social Responsiveness Scale were no longer significant. These findings demonstrate the need for studies to have adequate sample sizes to provide sufficient power to disentangle effects related specifically to tics rather than to other symptoms. Another study was designed to separate effects attributable to OCD54. Sustained attention, using a continuous performance test, was examined in 48 children and adolescents who had OCD alone, tic disorders (TD) alone or both OCD and TD. A high rate of ADHD was seen in all groups (62% of the OCD+TD group, 27% in the TD alone group, and 20% in the OCD alone group). Anxiety was also frequent (77% in the OCD+TD group compared to 49% for the other two groups combined). The OCD+ TD group had more errors of omission and higher reaction time variability. These results of this study provide additional evidence that the OCD+TD phenotype is associated with more severe symptoms including attentional difficulties and symptoms of anxiety.\n\nTwo studies examined motor control. In a clever analysis of video recordings of the eyes55, spontaneous blink rate, which is related to dopamine levels, was higher in children with TS than in controls both during task performance and during a rest period. In contrast, pupil diameter, which is related to norepinephrine levels, was correlated with anxiety in TS subjects although not in controls. Researchers also used a cognitive control task to measure ability to properly update current task information, ignore competing information when selecting between response options, and retrieve and use relevant response contingencies. Accuracy on this cognitive control task accounted for half of tic severity variance. In an unrelated study, TS children without ADHD or OCD had significantly greater difficulty maintaining postural stability than did age- and gender-matched controls, especially when subjects had access only to accurate vestibular, rather than visual or somatosensory, cues56.\n\nSomatosensory sensitivity was compared in adults with “pure” TS and controls by establishing thresholds for externally applied stimuli57. No differences were found between the two groups, supporting the view that the sensory abnormalities seen in TS may be related to abnormal interoceptive awareness or abnormal central sensorimotor processing.\n\nThree recent studies examined the effects of attention on tic frequency and the results have implications for how treatment protocols could be modified to increase effectiveness. In one study, the role of attention on tic frequency was examined under several conditions58. Tic frequencies were lower for 12 TS subjects during a baseline condition when they were alone in a room compared to when they were alone in a room looking at themselves in a mirror. Researchers then determined whether the increase in frequency was due to increased attention to the tics themselves or due to increased self-awareness in general. In addition to the conditions previously described, 16 subjects were also shown videos of themselves while they were not ticcing. Tic frequency was again lower during the baseline compared to the mirror condition. Tic frequency was even lower when subjects were watching the video of themselves while not ticcing. The authors suggest that future treatments teach patients to attend to states when they experience fewer tics. Another study of TS adults59 compared tic frequency while subjects were engaged in tasks that involved attending to particular fingers, colored circles, or whether a tic had occurred during a specific 2-second interval. Observations for these tasks were made both during free ticcing and tic suppression. Not surprisingly, more tics were seen during a baseline free ticcing condition. During the attention tasks, tic frequency was greatest while subjects focused on their tics. In contrast, tic frequency decreased during the color attention condition and decreased further during the finger attention condition. When subjects suppressed their tics, they reduced their baseline tic frequency similarly across all attention conditions. These results are consistent with the idea that internally-directed attention, especially with a focus on tics, may contribute to momentary increases in tic severity. The authors suggested that behavioral treatment might be more effective if it focused on teaching patients to focus on external events and voluntary actions when they are in situations that are most likely to result in ticcing. Anecdotal evidence has suggested that tics decrease when people are involved in musical activity, so Bodeck et al. systematically studied the effects of music.60 Questionnaires completed by 29 patients supported the idea that listening to music and performing music decrease tic frequency. Eight TS subjects were then observed in a variety of conditions. Tics were almost completely eliminated when subjects were performing music. Listening to music and using mental imagery of musical performance also resulted in decreases in tic frequency. The authors suggested that focused attention, along with fine motor control and goal-directed behavior, produced the decrease in tics.\n\nThe stereotyped nature of tics has led some to suggest that the neural systems involved in habitual behavior may also be associated with tic generation. A complex, three-stage instrumental learning paradigm was used to compare medicated and unmedicated TS adults with a control group to determine whether they differed in goal-directed vs. habitual behavior61. During the first stage, subjects learned to associate six different stimuli with six specific outcome pictures and a specific response (i.e., left or right key press). During the second stage, subjects were presented with two outcomes with an indication that one outcome was devalued (i.e., no longer associated with point rewards) and subjects had to press the key associated with the outcome that would still generate points. During the third stage (i.e., “slip-of-action” stage), the six outcomes were presented simultaneously with indications that two outcomes were devalued so that responding to the associated stimuli would no longer generate points. Subjects were instructed to press the key associated with stimuli associated with the still valued outcomes (i.e., “Go”) and withhold the response (i.e., “No-Go”) for stimuli associated with devalued outcomes. This task determined whether excessive “slips of action” were related to outcome devaluation insensitivity. A control Go/No-Go task, which involved devaluation of cueing stimuli, was used to measure to measure response rates where high rates on this task would indicate working memory deficits or deficient response inhibition. There were no group performance differences for the first two stages of the instrumental learning task or on the baseline Go/No-Go task. However, unmedicated patients showed a significantly higher response rate to devalued outcomes compared to controls (in Bonferroni-corrected post hoc analyses), while there was no difference between medicated subjects and controls. In addition, tic severity in unmedicated subjects was correlated with response rates to devalued outcomes and with stronger structural connectivity between the right supplementary motor cortex and the posterior putamen. The results obtained in this study contrasted with results on similar tasks obtained with subjects with obsessive-compulsive disorder without tics. The authors argued that over-reliance on habits in OCD without tics is associated with impaired knowledge of response-outcome associations, while this type of learning was intact in both TS groups in this study. They concluded that habit formation is enhanced in unmedicated TS subjects but medication may normalize responses.\n\nPsychotherapy. 2015 saw several practical advances in the psychotherapeutic treatments available for tics. TicHelper.com is a commercial adaptation of Comprehensive Behavioral Intervention for Tics (CBIT) to the Internet, discussed at the London congress in 2015.62 It is potentially an important treatment option, especially for the many TS patients who do not live near a behavior therapist. Efficacy testing is ongoing (see the trial summary at ClinicalTrials.gov).\n\nMcGuire et al. dug into the data from 2 previously reported, pivotal, randomized controlled CBIT studies that together enrolled over 200 children and adults63. The superior treatment benefit from CBIT, compared to a control therapy, could be attributed to differential improvement in only a few types of tics, including throat clearing, sniffing, and complex tics. In general, vocal tics were more likely to improve following CBIT treatment. The controlled breathing used as a competing response for vocal tics may have allowed patients to direct attention away from the associated premonitory urges in a way that muscle-tensing competing responses for motor tics did not. This report also extends previous information about premonitory phenomena, including varying prevalence of premonitory urges across specific tic types.\n\n“Living with Tics” is a modularized cognitive-behavioral treatment focused on decreasing tic-related impairment and improving quality of life. This treatment program was recently tested in a randomized, wait-list control study, with the active intervention including up to 10 weekly sessions for children and adolescents64. Treatment modules focused on a variety of themes including self-esteem, emotion regulation, parent training, cognitive restructuring, coping at school, overcoming tic-related avoidance, and 1 or 2 sessions of habit-reversal training. Active treatment led to improved child-rated quality of life and reduced blinded clinician-rated tic impairment compared to the wait-list control group. An additional 7 wait-listed youth then participated in the treatment program resulting in data on 19 participants for open-trial analyses. With the larger sample size the reductions in tic severity (i.e., 30%), anxiety, obsessive-compulsive symptoms, and parent-rated impairment were significant. Both the youth and their parents reported satisfaction with the intervention, serving as a reminder that improving quality of life can be a desired treatment goal.\n\nA small, uncontrolled open trial of mindfulness-based stress reduction treatment involved 18 individuals who were at least 16 years old.65 Treatment consisted of eight weekly two-hour group classes and one four-hour retreat. Participants were taught a tic-specific meditation exercise that involved noticing any urges to tic while maintaining a focus on one’s breathing rather than trying to change or eliminate the urge to tic. Only one subject dropped out and overall tic severity was decreased by 20% for participants who completed the program. An independent evaluator rated ten subjects as \"much improved\" or “very much improved” and these subjects were considered treatment responders. The gains for the fifteen participants who had not had medication changes in the interim were maintained at a one month follow-up visit.\n\nAn unblinded trial of psychotherapy using a cognitive psychophysiological model of tic behavior was conducted with 102 adults who had TS or chronic tic disorder66. Ten weeks of individual psychotherapy involved a number of components including increasing tic awareness; improving muscle control; preventing excessive muscle tension; decreasing an overactive action style; identifying low and high risk activities in terms of tic probability; highlighting differences in behaviors, thoughts and feelings related to differences in tic probability; decreasing perfectionistic beliefs linked to tension; generalization; and relapse prevention. These psychotherapy components were chosen because prior research suggested that some people with tics have perfectionist beliefs leading to an “impulsive overactive style” that produces frustration and tension in addition to tics. Large effect sizes were seen for both patient groups compared to the waiting list control group: 65% of the chronic tic disorder group and 74% of the TS group had reductions of more than 35% on the Tourette Syndrome Global Scale (TSGS). YGTSS total tic scores were also significantly decreased for both patient groups. At the end of treatment 78 of 85 completers were rated as having no more than mild symptoms, regardless of the starting severity, while the other 7 were considered to have moderate symptoms. Large effect sizes were seen for tic subtypes (i.e., simple, complex, motor, phonic) and similar results were seen across tic body locations. The TSGS improvements were maintained in the 52 subjects who completed a 6-month follow-up evaluation.\n\nA study by the same group examined the effects of this psychotherapy approach on event-related potentials and lateralized readiness potentials (LRPs)67. EEGs were recorded for 20 TS subjects and 20 control subjects matched for age, sex and intelligence while performing a stimulus-response compatibility inhibition task. During the No-Go condition, the TS group exhibited a delayed and overactivated frontal late positive component on the No-Go portion of the task. The authors considered this frontal activation as evidence of an adaptive mechanism that allowed patients to perform similarly to the control subjects on the task. This difference did not normalize with psychotherapy, consistent with this interpretation. TS subjects exhibited a larger incorrect activation of the stimulus-locked LRP (sLRP) in addition to larger correct activation of the sLRP and delayed correct activation sLRP onset. Although sLRP onset and response-locked LRP (rLRP) peak normalized after psychotherapy, the larger sLRP amplitude did not. The authors suggested that therapy produced some normalization of activation in the premotor and motor cortex.\n\nMedication. Efforts to maximize the value of pharmacological treatment continue. Lemmon et al. reported on a carefully designed, thoughtful pilot study of glutamatergic modulators as tic treatment68. Twenty-three children with TS completed a double-blind, parallel group study involving 6 weeks of placebo, D-serine (up to 30 mg/kg/day) or riluzole (up to 200 mg/day). YGTSS total tic scores improved by 25–38% in each group, without significant group differences. Although power was limited by the small sample size, this null result argues against eager pursuit of glutamatergic medications for TS at this juncture.\n\nA meta-analysis of 22 randomized, controlled trials (RCTs) involving 2,385 children with ADHD found no causal relationship between stimulants and onset of tics69. Rather, tics were associated with ADHD itself (5.7% in the psychostimulant groups and 6.5% in the placebo groups). This summary of previous evidence hopefully can further reassure patients and prescribers that stimulants do not cause tics. The evidence for this conclusion is strongest for methylphenidate (19 of the 22 trials), and in a large RCT in children with TS and ADHD, tics improved significantly with methylphenidate70. In the meta-analysis, results were similar for the 3 trials involving amphetamines, but tic severity did increase in 12 adults with TS after a single intravenous dose of 0.3mg/kg D-amphetamine71.\n\nFour patients with treatment-refractory TS were studied in an early report of results using vigabatrin, a medication from the GABA-aminotransferase inhibitor class72. One patient had a clinically significant reduction in tics while two others had tic reduction of approximately 25% but did not report subjective clinical improvement. Further study will be needed.\n\nNeurosurgery. A revised consensus statement on the use of DBS in TS appeared recently73. It provides an important update to the 2006 recommendations74, guided by almost a decade of generally positive results from an increasingly varied set of patients, though with limited evidence from randomized allocation treatment studies. Several major changes were made in the recommendations. First, the recommendation that DBS patients be 25 years or older has been replaced by a focus on clinical symptoms and severity. However, at a minimum, local ethics committee involvement was recommended for patients under 18 years of age. The second major modification in the guidelines was that a patient have a caregiver who would be available to accompany the patient to frequent follow-up appointments. In addition, the group recommended absence of active suicidal or homicidal ideation for 6 months prior to surgery. The final change in the guidelines recommended identifying and addressing personality disorders, malingering, factitious symptoms, embellishment, and other factors that can substantially complicate assessment and treatment.\n\nTreatment effectiveness was examined by randomly allocated to DBS on-stimulation or off-stimulation conditions for the first 3 months after DBS leads were placed in the GPi (globus pallidus, pars interna). Patients were then switched to the other condition for another 3 months75. Ratings were collected blind to stimulation status, and 13 patients completed assessments during both conditions. Total YGTSS scores were 15% lower at the end of the on-stimulation period compared with the off-stimulation period (p=0.048). Three serious adverse events occurred: two infections in the DBS hardware and one episode of hypomania. Further improvements were seen during the long-term open label treatment period. This study is important as it provides evidence of efficacy of stimulation per se from a study with random assignment to initial condition.\n\nA case study raised the issue of temporary DBS treatment in a patient with TS and ADHD treated with thalamic DBS from age of 17 to 2376. This case highlights the issue of DBS in minors discussed in the consensus statement73, but without adequate controls it is impossible to know whether DBS was the cause of tic improvement.\n\nOther treatment. Lisanby and colleagues reported a careful, randomized controlled trial of repetitive transcranial magnetic stimulation (rTMS) aimed at the supplementary motor area (SMA) in 20 adults with TS77. There were suggestions of improvement, but on average, patients in the sham treatment group improved almost as much as those in the active treatment group. However, rTMS most effectively stimulates superficial regions of cortex, and an Israeli group employed a new coil designed to provide deeper stimulation in an open-label study of 12 patients with treatment-refractory TS78. On average, the 12 patients did not improve significantly, but the treatment was well tolerated, and a post hoc analysis showed benefit in the 6 patients who also had OCD. A double blind, sham-controlled study in TS patients with OCD will be needed to confirm efficacy.\n\nCranial electrical stimulation (CES) was used in an open label trial to treat 42 children with TS who were less than 12 years old79. Patients applied electrodes to their earlobes when they went to bed so that they could receive the treatment on a daily basis for 24 weeks. Treatment sessions lasted 60 minutes and children were allowed to sleep during treatment. Only one child dropped out before completing the study. The mean YGTSS score significantly decreased from 26 when they were initially seen to 11 after treatment completion. These results must be viewed as preliminary, since an RCT is required to rule out spontaneous improvement.\n\nAnother treatment undergoing initial efficacy testing is an oral orthotic device with which some patients have reported success. The rationale and design of the study, and initial safety results, were reported at the London meeting80, and efficacy testing continues (see the trial summary at ClinicalTrials.gov).\n\nTwo studies examined the association between parenting a TS child and parental stress. Using self-report data, Stewart and colleagues revealed that stress in parents of 74 children with Tourette syndrome was highly correlated with current ADHD symptom burden in the child (r = .57) and, less strongly, with the child’s OCD symptoms (r = .40)81. However, perhaps surprising some readers, parental stress was not significantly correlated with current tic severity (r = .18, p = .13). The correlation of parental stress with ADHD and OCD symptom severity was present to a similar degree in 48 control children without tics. These observations reinforce the need for optimal clinical management of TS to include appropriate treatment of associated comorbidities. In another study 28 parents of 21 TS children completed questionnaires on their anxiety, depression, and perceived stress82. Eight of the parents reported a moderate or severe level of anxiety. In general, parents varied in terms of their self-reports of their coping skills and social support availability. These results remind clinicians that it is important to evaluate the social support available to families of TS children, and suggest that increasing parental coping skills may improve the quality of life in TS families.\n\nAttitudes toward treatment were investigated using a survey of 295 parents of youth with TS and interviews with 42 young people with TS83. These subjects were identified through Tourettes Action, a non-profit organization in the U.K. Patients tended to view aripiprazole more positively than other treatments. Three quarters of parents wanted access to behavior therapy for their child, but reported substantial trouble finding it. On the other hand, many young people were skeptical that behavior therapy would help. Importantly, patients identified several treatment outcomes other than tics per se as important to them: reduction of premonitory urges, increasing control over tics, and reducing anxiety.\n\nDavid Shprecher and colleagues recently published a thoughtful, forward-looking, but concise review of Tourette research and treatment84. Mary Robertson85–88 reviews work that she and colleagues have done during her 35-year career treating TS patients and her 30 years of publishing TS research. Jankovic provided an overview of tic treatments89, and one of our colleagues highly recommended a review of current medication treatment practice in Germany90.\n\nJackson et al. describe how increased motor control occurs in TS adolescents along with a variety of compensatory mechanisms91. These researchers make the case that increased inhibitory capacity within higher-order motor regions, such as the SMA, may alter motor excitability. Hawksley et al. summarize evidence suggesting that autonomic dysfunction may have a role in both epilepsy and Tourette Syndrome92. They also discuss how electrodermal activity biofeedback reduced seizure activity significantly in patients with epilepsy. In contrast, TS subjects had difficulty reducing sympathetic activity, suggesting that treatment protocol modifications are needed for TS.\n\nIn December, 2015, a special issue on Tourette Syndrome appeared in the new journal Current Developmental Disorders Reports. Four of the 5 articles are clinically focused 13,93–95. The other article summarizes some of the recent advances in TS neuroimaging96.\n\nAbstracts of presentations from the 1st World Congress on Tourette Syndrome & Tic Disorders, London, 24–26 June 2015, are available online. The Tourette Association at regular intervals posts links on its web page to selected scientific articles relevant to TS97. Finally, the authors are building the 2016 version of this article here, where readers are invited to submit suggestions or comments.\n\n\nConclusions\n\nTS is a complex condition to study because of the high rate of comorbidities, the differences in tic severity between people identified through community studies and those who participate in research studies conducted at tertiary care centers, the differences in types of tics (e.g., vocal vs. motor tics, simple tics vs. complex tics), childhood onset, and prominent fluctuations in severity over time. The studies published in 2015 demonstrate continued progress in addressing some of these obstacles.\n\nAlthough many early research studies focused on the roles of dopamine and the basal ganglia in tic generation, more recent studies are adding to our understanding of the complexity of tic generation and identifying possible roles for glutamate and histamine, along with a variety of subcortical and cortical regions. A number of carefully executed, large treatment studies have been reported, yet none is effective for more than about half of the patients enrolled, and side effects, cost or availability limit many existing treatments. Novel treatments are still needed.",
"appendix": "Author contributions\n\n\n\nBoth authors contributed to all phases of this work and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nKJB participated in clinical trials supported by Psyadon Pharmaceuticals, Neurocrine Biosciences, Inc., and Acadia Pharmaceuticals, and was compensated by Acadia for advisory board and speakers bureau participation.\n\n\nGrant information\n\nThis work was supported in part by U.S. National Institutes of Health (NIH) grants K24 MH087913 and R21 NS091635, by a research grant from the Tourette Association of America to author CAR, and by the McDonnell Center for Systems Neuroscience at Washington University.\n\nThe authors confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The opinions expressed are the authors’ and may not reflect the opinion of the funders.\n\n\nReferences\n\nRichards CA, Black KJ: Tourette Syndrome research highlights 2014 [version 2; referees: 1 approved, 2 approved with reservations]. F1000Res. 2015; 4: 69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlack KJ: F1000Research: Tics welcomes you to 21st century biomedical publishing [version 1; referees: not peer reviewed]. F1000Res. 2014; 3: 272. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGanos C, Bongert J, Asmuss L, et al.: The somatotopy of tic inhibition: Where and how much? Mov Disord. 2015; 30(9): 1184–9. PubMed Abstract | Publisher Full Text\n\nGreene DJ, Koller JM, Robichaux-Viehoever A, et al.: Reward enhances tic suppression in children within months of tic disorder onset. Dev Cogn Neurosci. 2015; 11: 65–74. The ability to suppress tics is present within months of tic onset. Immediate, contingent reward enhances these children’s ability to suppress tics. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsuda N, Kono T, Nonaka M, et al.: Self-initiated coping with Tourette’s syndrome: Effect of tic suppression on QOL. Brain Dev. 2016; 38(2): 233–41. PubMed Abstract | Publisher Full Text\n\nKano Y, Matsuda N, Nonaka M, et al.: Sensory phenomena related to tics, obsessive-compulsive symptoms, and global functioning in Tourette syndrome. Compr Psychiatry. 2015; 62: 141–146. PubMed Abstract | Publisher Full Text\n\nGanos C, Garrido A, Navalpotro-Gómez I, et al.: Premonitory urge to tic in Tourette’s is associated with interoceptive awareness. Mov Disord. 2015; 30(9): 1198–202. PubMed Abstract | Publisher Full Text\n\nWong LC, Huang HL, Weng WC, et al.: Increased risk of epilepsy in children with Tourette syndrome: A population-based case-control study. Res Dev Disabil. 2016; 51–52: 181–7. PubMed Abstract | Publisher Full Text\n\nHirschtritt ME, Lee PC, Pauls DL, et al.: Lifetime prevalence, age of risk, and genetic relationships of comorbid psychiatric disorders in Tourette syndrome. JAMA Psychiatry. 2015; 72(4): 325–333. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEapen V, Robertson MM: Are there distinct subtypes in Tourette syndrome? Pure-Tourette syndrome versus Tourette syndrome-plus, and simple versus complex tics. Neuropsychiatr Dis Treat. 2015; 11: 1431–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCox JH, Cavanna AE: Irritability symptoms in Gilles de la Tourette syndrome. J Neuropsychiatry Clin Neurosci. 2015; 27(1): 42–7. PubMed Abstract | Publisher Full Text\n\nEddy CM, Macerollo A, Martino D, et al.: Interpersonal reactivity differences in Tourette syndrome. Psychiatry Res. 2015; 228(3): 932–5. PubMed Abstract | Publisher Full Text\n\nBudman CL, Rosen M, Shad S: Fits Tantrums, and Rages in TS and Related Disorders. Curr Dev Disord Rep. 2015; 2(4): 273–284. Publisher Full Text\n\nBalottin L, Selvini C, Luoni C, et al.: Personality Profile of Male Adolescents With Tourette Syndrome: A Controlled Study. J Child Neurol. 2016; 31(3): 294–9. PubMed Abstract | Publisher Full Text\n\nTrillini MO, Müller-Vahl KR: Patients with Gilles de la Tourette syndrome have widespread personality differences. Psychiatry Res. 2015; 228(3): 765–73. PubMed Abstract | Publisher Full Text\n\nZappella M, Einspieler C, Bartl-Pokorny KD, et al.: What do home videos tell us about early motor and socio-communicative behaviours in children with autistic features during the second year of life--An exploratory study. Early Hum Dev. 2015; 91(10): 569–75. PubMed Abstract | Publisher Full Text\n\nRicher F, Daghfal R, Rouleau GA, et al.: Clinical features associated with an early onset in chronic tic disorders. Psychiatry Res. 2015; 230(3): 745–8. PubMed Abstract | Publisher Full Text\n\nLichter DG, Finnegan SG: Influence of gender on Tourette syndrome beyond adolescence. Eur Psychiatry. 2015; 30(2): 334–340. PubMed Abstract | Publisher Full Text\n\nHuang A, Yu D, Davis L, et al.: A large-scale study of rare copy number variation identifies novel loci associated with Tourette’s syndrome. 2015. http://eventmobi.com/api/events/7343/documents/download/40ae88af-e882-4b97-a43a-b361d8ff3d67.pdf/as/Aarchived by WebCite®. Reference Source\n\nClarke RA, Eapen V: Balance within the Neurexin Trans-Synaptic Connexus Stabilizes Behavioral Control. Front Hum Neurosci. 2014; 8: 52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYe H, Tan YL, Ponniah S, et al.: Neural recognition molecules CHL1 and NB-3 regulate apical dendrite orientation in the neocortex via PTP alpha. EMBO J. 2008; 27(1): 188–200. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTakeda Y, Akasaka K, Lee S, et al.: Impaired motor coordination in mice lacking neural recognition molecule NB-3 of the contactin/F3 subgroup. J Neurobiol. 2003; 56(3): 252–65. PubMed Abstract | Publisher Full Text\n\nHu J, Liao J, Sathanoori M, et al.: CNTN6 copy number variations in 14 patients: a possible candidate gene for neurodevelopmental and neuropsychiatric disorders. J Neurodev Disord. 2015; 7(1): 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBertelsen B, Stefánsson H, Riff Jensen L, et al.: Association of AADAC Deletion and Gilles de la Tourette Syndrome in a Large European Cohort. Biol Psychiatry. 2016; 79(5): 383–391. PubMed Abstract | Publisher Full Text\n\nde Leeuw C, Goudriaan A, Smit AB, et al.: Involvement of astrocyte metabolic coupling in Tourette syndrome pathogenesis. Eur J Hum Genet. 2015; 23(11): 1519–1522. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu D, Mathews CA, Scharf JM, et al.: Cross-disorder genome-wide analyses suggest a complex genetic relationship between Tourette’s syndrome and OCD. Am J Psychiatry. 2015; 172(1): 82–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Haan MJ, Delucchi KL, Mathews CM, et al.: Tic symptom dimensions and their heritabilities in Tourette’s syndrome. Psychiatr Genet. 2015; 25(3): 112–8. PubMed Abstract | Publisher Full Text\n\nBen-Shlomo Y, Scharf JM, Miller LL, et al.: Parental mood during pregnancy and post-natally is associated with offspring risk of Tourette syndrome or chronic tics: prospective data from the Avon Longitudinal Study of Parents and Children (ALSPAC). European Child & Adolescent Psychiatry. 2016; 25(4): 373–381. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLennington JB, Coppola G, Kataoka-Sasaki Y, et al.: Transcriptome analysis of the human striatum in Tourette Syndrome. Biol Psychiatry. 2016; 79(5): 372–382. pii: S0006-3223(14)00551-4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu M, Kobets A, Du JC, et al.: Targeted ablation of cholinergic interneurons in the dorsolateral striatum produces behavioral manifestations of Tourette syndrome. Proc Natl Acad Sci U S A. 2015; 112(3): 893–898. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIsraelashvili M, Bar-Gad I: Corticostriatal Divergent Function in Determining the Temporal and Spatial Properties of Motor Tics. J Neurosci. 2015; 35(50): 16340–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPogorelov V, Xu M, Smith HR, et al.: Corticostriatal interactions in the generation of tic-like behaviors after local striatal disinhibition. Exp Neurol. 2015; 265: 122–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNordstrom EJ, Bittner KC, McGrath MJ, et al.: \"Hyperglutamatergic cortico-striato-thalamo-cortical circuit\" breaker drugs alleviate tics in a transgenic circuit model of Tourette׳s syndrome. Brain Res. 2015; 1629: 38–53. PubMed Abstract | Publisher Full Text\n\nGreene DJ, Black KJ, Schlaggar BL: Considerations for MRI study design and implementation in pediatric and clinical populations. Dev Cogn Neurosci. 2016; 18: 101–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReuter M, Tisdall MD, Qureshi A, et al.: Head motion during MRI acquisition reduces gray matter volume and thickness estimates. Neuroimage. 2015; 107: 107–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTisdall MD, Reuter M, Qureshi A, et al.: Prospective motion correction with volumetric navigators (vNavs) reduces the bias and variance in brain morphometry induced by subject motion. Neuroimage. 2016; 127: 11–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDraper A, Jackson GM, Morgan PS, et al.: Premonitory urges are associated with decreased grey matter thickness within the insula and senso-rimotor cortex in young people with Tourette syndrome. J Neuropsychol. 2016; 10(1): 143–53. PubMed Abstract | Publisher Full Text\n\nTinaz S, Malone P, Hallett M, et al.: Role of the right dorsal anterior insula in the urge to tic in Tourette syndrome. Mov Disord. 2015; 30(9): 1190–7. PubMed Abstract | Publisher Full Text\n\nPower JD, Schlaggar BL, Petersen SE: Recent progress and outstanding issues in motion correction in resting state fMRI. Neuroimage. 2015; 105: 536–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZapparoli L, Porta M, Paulesu E: The anarchic brain in action: the contribution of task-based fMRI studies to the understanding of Gilles de la Tourette syndrome. Curr Opin Neurol. 2015; 28(6): 604–11. PubMed Abstract | Publisher Full Text\n\nZapparoli L, Porta M, Gandola M, et al.: A functional magnetic resonance imaging investigation of motor control in Gilles de la Tourette Syndrome during imagined and executed movements. Eur J Neurosci. 2016; 43(4): 494–508. PubMed Abstract | Publisher Full Text\n\nWilliams AC III, Greene DJ, Koller JM, et al.: A pilot multicenter study of brain structure in pediatric Tourette syndrome. 2015; http://eventmobi.com/tourette2015/documents/112809 archived by WebCite®. Reference Source\n\nWolff N, Luehr I, Sender J, et al.: A DTI study on the corpus callosum of treatment-naïve boys with ’pure’ Tourette syndrome. Psychiatry Res. 2016; 247: 1–8. PubMed Abstract | Publisher Full Text\n\nMorand-Beaulieu S, Fernandes I, Sauvé G, et al.: Electrophysiological similarities in Tourette syndrome and chronic tic disorder. Int J Dev Neurosci. 2015; 47(Part A): 63. Publisher Full Text\n\nGermain M, Chuang MN, Sauvé G, et al.: The impact of comorbidity on event-related brain topography in patients presenting chronic tic disorders and Tourette syndrome. Int J Dev Neurosci. 2015; 47(Part A): 102. Publisher Full Text\n\nBour LJ, Ackermans L, Foncke EM, et al.: Tic related local field potentials in the thalamus and the effect of deep brain stimulation in Tourette syndrome: Report of three cases. Clin Neurophysiol. 2015; 126(8): 1578–88. PubMed Abstract | Publisher Full Text\n\nPuts NA, Harris AD, Crocetti D, et al.: Reduced GABAergic inhibition and abnormal sensory symptoms in children with Tourette syndrome. J Neurophysiol. 2015; 114(2): 808–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbi-Jaoude E, Segura B, Obeso I, et al.: Similar striatal D2/D3 dopamine receptor availability in adults with Tourette syndrome compared with healthy controls: A [11C]-(+)-PHNO and [11C]raclopride positron emission tomography imaging study. Hum Brain Mapp. 2015; 36(7): 2592–601. PubMed Abstract | Publisher Full Text\n\nRobertson CL, Ishibashi K, Mandelkern MA, et al.: Striatal D1- and D2-type dopamine receptors are linked to motor response inhibition in human subjects. J Neurosci. 2015; 35(15): 5990–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCox SM, Frank MJ, Larcher K, et al.: Striatal D1 and D2 signaling differentially predict learning from positive and negative outcomes. Neuroimage. 2015; 109: 95–101. PubMed Abstract | Publisher Full Text\n\nBolam JP, Ellender TJ: Histamine and the striatum. Neuropharmacology. 2015; pii: S0028-3908(15)30059-9. PubMed Abstract | Publisher Full Text\n\nEddy CM, Cavanna AE: Triangles, tricks and tics: Hyper-mentalizing in response to animated shapes in Tourette syndrome. Cortex. 2015; 71: 68–75. PubMed Abstract | Publisher Full Text\n\nGüler AS, Berkem M, Yazgan Y, et al.: Cognitive Flexibility and Social Responsiveness in Children and Adolescents with Tourette Syndrome. Child Psychiatry Hum Dev. 2015; 46(6): 940–50. PubMed Abstract | Publisher Full Text\n\nLucke IM, Lin C, Conteh F, et al.: Continuous performance test in pediatric obsessive-compulsive disorder and tic disorders: the role of sustained attention. CNS Spectr. 2015; 20(5): 479–89. PubMed Abstract | Publisher Full Text\n\nTharp JA, Wendelken C, Mathews CA, et al.: Tourette Syndrome: Complementary Insights from Measures of Cognitive Control, Eyeblink Rate, and Pupil Diameter. Front Psychiatry. 2015; 6: 95. PubMed Abstract | Publisher Full Text\n\nLiu WY, Ya-TingHsu, Lien HY, et al.: Deficits in sensory organization for postural stability in children with Tourette syndrome. Clin Neurol Neurosurg. 2015; 129(Suppl 1): S36–40. PubMed Abstract | Publisher Full Text\n\nSchunke O, Grashorn W, Kahl U, et al.: Quantitative Sensory Testing in adults with Tourette syndrome. Parkinsonism Relat Disord. 2016; 24: 132–6. PubMed Abstract | Publisher Full Text\n\nBrandt VC, Lynn MT, Obst M, et al.: Visual feedback of own tics increases tic frequency in patients with Tourette’s syndrome. Cogn Neurosci. 2015; 6(1): 1–7. PubMed Abstract | Publisher Full Text\n\nMisirlisoy E, Brandt V, Ganos C, et al.: The relation between attention and tic generation in Tourette syndrome. Neuropsychology. 2015; 29(4): 658–65. PubMed Abstract | Publisher Full Text\n\nBodeck S, Lappe C, Evers S: Tic-reducing effects of music in patients with Tourette’s syndrome: Self-reported and objective analysis. J Neurol Sci. 2015; 352(1–2): 41–47. PubMed Abstract | Publisher Full Text\n\nDelorme C, Salvador A, Valabrègue R, et al.: Enhanced habit formation in Gilles de la Tourette syndrome. Brain. 2016; 139(pt 2): 605–15. PubMed Abstract | Publisher Full Text\n\nHimle M, Mouton-Odum S, Reamer R, et al.: Development and Initial Feasibility Testing of TicHelper: A Self-Administered Interactive Program for Teaching Comprehensive Behavioral Intervention for Tics. 2015. Reference Source\n\nMcGuire JF, Piacentini J, Scahill L, et al.: Bothersome tics in patients with chronic tic disorders: Characteristics and individualized treatment response to behavior therapy. Behav Res Ther. 2015; 70: 56–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcGuire JF, Arnold E, Park JM, et al.: Living with tics: Reduced impairment and improved quality of life for youth with chronic tic disorders. Psychiatry Res. 2015; 225(3): 571–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReese HE, Vallejo Z, Rasmussen J, et al.: Mindfulness-based stress reduction for Tourette Syndrome and Chronic Tic Disorder: a pilot study. J Psychosom Res. 2015; 78(3): 293–298. PubMed Abstract | Publisher Full Text\n\nO’Connor K, Lavoie M, Blanchet P, et al.: Evaluation of a cognitive psychophysiological model for management of tic disorders: an open trial. Br J Psychiatry. 2015; pii: bjp.bp.114.154518. PubMed Abstract | Publisher Full Text\n\nMorand-Beaulieu S, O’Connor KP, Sauvé G, et al.: Cognitive-behavioral therapy induces sensorimotor and specific electrocortical changes in chronic tic and Tourette’s disorder. Neuropsychologia. 2015; 79(pt B): 310–21. PubMed Abstract | Publisher Full Text\n\nLemmon ME, Grados M, Kline T, et al.: Efficacy of Glutamate Modulators in Tic Suppression: A Double-Blind, Randomized Control Trial of D-serine and Riluzole in Tourette Syndrome. Pediatr Neurol. 2015; 52(6): 629–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCohen SC, Mulqueen JM, Ferracioli-Oda E, et al.: Meta-Analysis: Risk of Tics Associated With Psychostimulant Use in Randomized, Placebo-Controlled Trials. J Am Acad Child Adolesc Psychiatry. 2015; 54(9): 728–36. PubMed Abstract | Publisher Full Text\n\nTourette’s Syndrome Study Group: Treatment of ADHD in children with tics: a randomized controlled trial. Neurology. 2002; 58(4): 527–36. PubMed Abstract | Publisher Full Text\n\nDenys D, de Vries F, Cath D, et al.: Dopaminergic activity in Tourette syndrome and obsessive-compulsive disorder. Eur Neuropsychopharmacol. 2013; 23(11): 1423–31. PubMed Abstract | Publisher Full Text\n\nCatalyst Pharmaceutical Partners Inc.: Catalyst Pharmaceuticals Announces Encouraging Top-Line Results in Proof-of-Concept Trial of Vigabatrin in Patients With Treatment-Refractory Tourette's Disorder. 2015; http://ir.catalystpharma.com/releasedetail.cfm?ReleaseID=919254 archived by WebCite®, Accessed: 2015-09-05. Reference Source\n\nSchrock LE, Mink JW, Woods DW, et al.: Tourette syndrome deep brain stimulation: a review and updated recommendations. Mov Disord. 2015; 30(4): 448–471. PubMed Abstract | Publisher Full Text\n\nMink JW, Walkup J, Frey KA, et al.: Patient selection and assessment recommendations for deep brain stimulation in Tourette syndrome. Mov Disord. 2006; 21(11): 1831–8. PubMed Abstract | Publisher Full Text\n\nKefalopoulou Z, Zrinzo L, Jahanshahi M, et al.: Bilateral globus pallidus stimulation for severe Tourette’s syndrome: a double-blind, randomised crossover trial. Lancet Neurol. 2015; 14(6): 595–605. PubMed Abstract | Publisher Full Text\n\nZekaj E, Saleh C, Porta M, et al.: Temporary deep brain stimulation in Gilles de la Tourette syndrome: A feasible approach? Surg Neurol Int. 2015; 6: 122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLanderos-Weisenberger A, Mantovani A, Motlagh MG, et al.: Randomized Sham Controlled Double-blind Trial of Repetitive Transcranial Magnetic Stimulation for Adults With Severe Tourette Syndrome. Brain Stimul. 2015; 8(3): 574–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBloch Y, Arad S, Levkovitz Y: Deep TMS add-on treatment for intractable Tourette syndrome: A feasibility study. World J Biol Psychiatry. 2014; 1–5. in press. PubMed Abstract | Publisher Full Text\n\nQiao J, Weng S, Wang P, et al.: Normalization of Intrinsic Neural Circuits Governing Tourette’s Syndrome Using Cranial Electrotherapy Stimulation. IEEE Trans Biomed Eng. 2015; 62(5 ): 1272–80. PubMed Abstract | Publisher Full Text\n\nBennett S, Walkup J, Hindin J, et al.: Proof of Concept Study of an Oral Orthotic in Reducing Tic Severity in Youth with Chronic Tic Disorder and Tourette Syndrome. 2015; http://eventmobi.com/api/events/7343/documents/download/178e10b8-091c-411f-b9dc-97d62a321f39.pdf/as/Proofarchived by WebCite®. Reference Source\n\nStewart SB, Greene DJ, Lessov-Schlaggar CN, et al.: Clinical correlates of parenting stress in children with Tourette syndrome and in typically developing children. J Pediatr. 2015; 166(5): 1297–1302.e3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoussé V, Czernecki V, Denis P, et al.: Impact of Perceived Stress, Anxiety-Depression and Social Support on Coping Strategies of Parents Having A Child With Gilles de la Tourette Syndrome. Arch Psychiatr Nurs. 2016; 30(1): 109–13. PubMed Abstract | Publisher Full Text\n\nCuenca J, Glazebrook C, Kendall T, et al.: Perceptions of treatment for tics among young people with Tourette syndrome and their parents: a mixed methods study. BMC Psychiatry. 2015; 15(1): 46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShprecher DR, Kious BM, Himle MH: Advances in mechanistic understanding and treatment approaches to Tourette syndrome. Discov Med. 2015; 20(111): 295–301. PubMed Abstract\n\nRobertson MM: A personal 35 year perspective on Gilles de la Tourette syndrome: prevalence, phenomenology, comorbidities, and coexistent psychopathologies. Lancet Psychiatry. 2015; 2(1): 68–87. PubMed Abstract | Publisher Full Text\n\nRobertson MM: A personal 35 year perspective on Gilles de la Tourette syndrome: assessment, investigations, and management. Lancet Psychiatry. 2015; 2(1): 88–104. PubMed Abstract | Publisher Full Text\n\nRobertson MM: Corrections. A personal 35 year perspective on Gilles de la Tourette syndrome: prevalence, phenomenology, comorbidities, and coexistent psychopathologies. Lancet Psychiatry. 2015; 2(4): 291. PubMed Abstract | Publisher Full Text\n\nRobertson MM: Corrections. A personal 35 year perspective on Gilles de la Tourette syndrome: assessment, investigations, and management. Lancet Psychiatry. 2015; 2(4): 291. PubMed Abstract | Publisher Full Text\n\nJankovic J: Therapeutic Developments for Tics and Myoclonus. Mov Disord. 2015; 30(11): 1566–1573. PubMed Abstract | Publisher Full Text\n\nBachmann CJ, Roessner V, Glaeske G, et al.: Trends in psychopharmacologic treatment of tic disorders in children and adolescents in Germany. Eur Child Adolesc Psychiatry. 2015; 24(2): 199–207. PubMed Abstract | Publisher Full Text\n\nJackson GM, Draper A, Dyke K, et al.: Inhibition, Disinhibition, and the Control of Action in Tourette Syndrome. Trends Cogn Sci. 2015; 19(11): 655–65. PubMed Abstract | Publisher Full Text\n\nHawksley J, Cavanna AE, Nagai Y: The role of the autonomic nervous system in Tourette Syndrome. Front Neurosci. 2015; 9: 117. PubMed Abstract | Publisher Full Text\n\nHanks CE, Lewin AB, Jane MP, et al.: Social Deficits and Autism Spectrum Disorders in Tourette’s Syndrome. Curr Dev Disord Rep. 2015; 2(4): 285–292. Publisher Full Text\n\nCoffey Barbara J: Complexities for Assessment and Treatment of Co-Occurring ADHD and Tics. Curr Dev Disord Rep. 2015; 2(4): 293–299. Publisher Full Text\n\nMcGuire JF, Ricketts EJ, Piacentini J, et al.: Behavior Therapy for Tic Disorders: an Evidenced-Based Review and New Directions for Treatment Research. Curr Dev Disord Rep. 2015; 2(4): 309–317. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreene DJ, Schlaggar BL, Black KJ: Neuroimaging in Tourette Syndrome: Research Highlights From 2014–2015. Curr Dev Disord Rep. 2015; 2(4): 300–308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTourette Association of America: Recently published scientific articles. 2015; Accessed: 2015-09-05. Reference Source"
}
|
[
{
"id": "14667",
"date": "06 Jul 2016",
"name": "Andreas Hartmann",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nA very comprehensive review of the TS literature in 2015 for which the authors must be thanked. Generally speaking, they provide a brief overview of the main results / points of each paper but do not dwell into interpretations and discussions, which is fine with me; all readers interested can look up the primary article. Also, there are no relevant articles missing from what I can gather as I have a weekly Pubcrawler alert for TS. The choice of papers is subjective, of course, and I was surprised not to find a few of our own group’s contributions, listed below, but I leave it up to the authors to elaborate on them or not.",
"responses": []
},
{
"id": "14670",
"date": "12 Jul 2016",
"name": "Keith A. Coffman",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI read, with great interest, the manuscript by Richards and Black. Overall, it was outstanding and provides an excellent, comprehensive review of the latest literature on Tourette Syndrome.\nI recommend that it be indexed, without revisions.",
"responses": []
},
{
"id": "14669",
"date": "12 Jul 2016",
"name": "Douglas W. Woods",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis was a well-written overview of research published in the TS field during 2015. I have no major concerns with the paper, but a couple of issues should be addressed in the paper.\nThe authors should make clear that this is not a methodological critique of any particular study or set of studies. It is simply an overview.\n\nThe authors should take care not to insert their subjective opinions into the paper. For example, on page 2, the authors note that \"The most important genetic news of the year may have been....\" I would like the readers to make that conclusion rather than the authors.\n\nThe authors conclusions about the effects of why vocal tics responded so well to CBIT are speculative (p. 8), and the authors should note that such an explanation would require confirmation.\n\nAt the end of the paper, I think the authors should make calls for additional studies. The 2015 year had things to say about why CBIT may or may not work, and the authors should call for additional work in that area. The authors should also call for large-scale translational studies that bring together various areas of basic science in the context of a clinical trial. This arrangement will lead to exciting new discoveries about tics and how treatments work. Finally, the authors should consider calling for research on optimizing treatment sequencing.\n\nOverall, this was a useful review.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1493
|
https://f1000research.com/articles/5-1478/v1
|
23 Jun 16
|
{
"type": "Review",
"title": "Recent advances in understanding transcription termination by RNA polymerase II",
"authors": [
"Travis J. Loya",
"Daniel Reines",
"Travis J. Loya"
],
"abstract": "Transcription termination is a fundamental process in which RNA polymerase ceases RNA chain extension and dissociates from the chromatin template, thereby defining the end of the transcription unit. Our understanding of the biological role and functional importance of termination by RNA polymerase II and the range of processes in which it is involved has grown significantly in recent years. A large set of nucleic acid-binding proteins and enzymes have been identified as part of the termination machinery. A greater appreciation for the coupling of termination to RNA processing and metabolism has been recognized. In addition to serving as an essential step at the end of the transcription cycle, termination is involved in the regulation of a broad range of cellular processes. More recently, a role for termination in pervasive transcription, non-coding RNA regulation, genetic stability, chromatin remodeling, the immune response, and disease has come to the fore. Interesting mechanistic questions remain, but the last several years have resulted in significant insights into termination and an increasing recognition of its biological importance.",
"keywords": [
"Transcription termination is the complex and tightly regulated process in which polymerase stops RNA chain elongation and dissociates from the end of transcription units. A multiplicity of termination factors",
"which assemble into a number of complexes",
"govern the biogenesis of various types of transcripts including messenger RNA (mRNA)",
"small nuclear RNA",
"small nucleolar RNA (snoRNA)",
"and long non-coding RNA (lncRNA). Much of what we know has been learned from studies in Saccharomyces cerevisiae",
"where one process operates to terminate short non-coding transcripts and another involves components of the polyadenylation (polyA) machinery and nucleases such as Xrn2 to terminate the synthesis of mRNA precursors. Recent reviews have described some of these fundamental mechanisms and factors involved in termination by RNA polymerase II (pol II)1–6. Here",
"we will focus on the role of the termination of transcription by pol II in a variety of biological contexts and describe how new discoveries have helped elucidate longstanding questions as well as contributed to the establishment of new paradigms. It has become clear that transcription termination occupies a critical role as a regulator of cellular processes. This process and the termination machinery occupy an increasingly important place in human health and disease."
],
"content": "Introduction\n\nTranscription termination is the complex and tightly regulated process in which polymerase stops RNA chain elongation and dissociates from the end of transcription units. A multiplicity of termination factors, which assemble into a number of complexes, govern the biogenesis of various types of transcripts including messenger RNA (mRNA), small nuclear RNA, small nucleolar RNA (snoRNA), and long non-coding RNA (lncRNA). Much of what we know has been learned from studies in Saccharomyces cerevisiae, where one process operates to terminate short non-coding transcripts and another involves components of the polyadenylation (polyA) machinery and nucleases such as Xrn2 to terminate the synthesis of mRNA precursors. Recent reviews have described some of these fundamental mechanisms and factors involved in termination by RNA polymerase II (pol II)1–6. Here, we will focus on the role of the termination of transcription by pol II in a variety of biological contexts and describe how new discoveries have helped elucidate longstanding questions as well as contributed to the establishment of new paradigms. It has become clear that transcription termination occupies a critical role as a regulator of cellular processes. This process and the termination machinery occupy an increasingly important place in human health and disease.\n\n\nEnd-of-open reading frame termination: recent studies on the torpedo and allosteric models of termination of mRNAs\n\nTwo prevailing hypotheses for how pol II terminates transcription have guided experiments for almost three decades; these are the so-called allosteric and torpedo models. In its simplest form, the first posits that as elongating pol II encounters a polyA signal, a physical change in the complex is triggered that provokes termination7. The torpedo model suggests more specifically that termination is facilitated by a nuclease brought to the transcript via the polyA signal and associated RNA processing machinery. After the co-transcriptional endonucleolytic cutting of the primary transcript in preparation for its polyadenylation, a nuclease engages the 5’ end of the 3’ portion of the RNA that remains polymerase associated, and digestion of this RNA ‘stump’ in the 5’ to 3’ direction enables the nuclease to chase down pol II, whereupon termination is triggered by an undetermined mechanism8. The difference between these two models may be reduced to the following question: what are the changes in pol II that persuade it to go from its default mode of continuing elongation into stopping and departing from the template? Although there has been some controversy over the relative acceptance of these two hypotheses, the two models are not mutually exclusive, and indeed a combination of both mechanisms has been proposed9–11.\n\nA recent test of the concept provides new insight showing that in vitro, a polyA signal is sufficient to induce elongation complex disassembly independent of transcript cleavage12. These authors propose that there is an important conformational shift that can be blocked by α-amanitin, implicating pol II itself in this transition. These findings seemingly support an allosteric component to the model, i.e. something happens to pol II following synthesis of the polyA signal in the nascent RNA, but cutting of the RNA is not needed. However, identifying that change in pol II has remained elusive. In an almost reciprocal study, it had been shown that exonucleolytic degradation of the transcript by the Rat1/Rai1 nuclease (the two subunits that compose yeast Xrn2) was found to be insufficient to dislodge pol II from its template in vitro10,13. So, even if a nuclease torpedo is involved in termination, it is not sufficient to complete the process.\n\nThe conserved Xrn2 nuclease has been considered a strong candidate to be the torpedo. Recently, Fong et al. showed that Xrn2 was necessary for proper termination at thousands of genes14. The requirement was not absolute in that loss of Xrn2 delayed, rather than prevented, termination, resulting in longer transcripts. This supports the idea of a kinetic competition in termination, as the Xrn2 torpedo chases the elongating polymerase while Xrn2 digests the transcript from 5’ to 3’. In cells with mutant pol II that was either abnormally slow or fast, it should take Xrn2 less and more time to catch pol II, respectively. Consistent with this prediction, slow pol II led to shorter terminated transcripts, and fast pol II produced longer terminated transcripts. This result can be taken as strong evidence in favor of the torpedo model. Curiously, Xrn2 was found to be important for termination, even for non-coding transcription units whose miRNA and lncRNA products are not acted upon by the cleavage and polyA machinery.\n\nIn another test of the need for cleavage of the primary transcript for termination, Schaughency et al. mapped the chromatin locations of pol II in a S. cerevisiae strain depleted for Ysh1, the endonuclease that generates the 3’ end that becomes polyadenylated11. The analysis revealed that pol II stops elongation but remains template bound, presumably because the lack of cleavage prevents a nuclease torpedo from gaining access to the elongation complex11. This gives rise to the idea that both an allosteric change resulting from crossing the polyA site and the torpedo effects of Rat1 are necessary for efficient pol II release in yeast.\n\nWhile implicated as an essential part of the torpedo model, there is also a growing body of research on Xrn2’s post-translational modification. Sansó et al.15 employed a chemical genetic screen introduced by Shokat and co-workers16,17 to look for substrates of the Cdk9 kinase, an enzyme implicated in transcription as a subunit of the pTEFb elongation factor18. By mutating its active site to accommodate a bulky ATP analog, these investigators could label kinase substrates specific for Cdk9 that could then be tracked by their covalently attached thiophosphate group. One such substrate was Xrn215. The phosphorylated form of Xrn2 could be found on chromatin and the modification was associated with a modest enhancement of its nuclease activity. Inhibition of Cdk9 kinase activity and mutation of the phosphorylated threonine in Xrn2 led to the expected defects in termination predicted by the torpedo model. This new finding incorporates Xrn2 into a cellular signaling pathway. It will be important to learn what cellular events activate and reverse phosphorylation-based control of the torpedo’s activity.\n\nXrn2 was also recently shown to regulate chromatin structure to promote meiotic gene silencing during vegetative growth in Schizosaccharomyces pombe19,20. In this case, heterochromatin formation was dependent on transcription and termination, which was coupled to the nucleolytic elimination of the resulting RNAs. Chalamcharla et al. showed that the Xrn2 homolog Dhp1 was necessary for premature termination of non-coding RNAs that marked sites of repression of facultative heterochromatin at meiotic genes19. Loss of dhp1 in an exosome-deficient strain resulted in compromised RNAi-mediated gene silencing, similar to the changes found in ago1 mutants. Tucker et al. also found a similar effect for Dhp1 in silencing and also showed that loss of Dhp1 resulted in defects in meiotic chromosome segregation20. Interestingly, the two groups have different mechanistic explanations for their observations. Chalamcharla et al. postulate that termination-coupled degradation by the exosome triggers recruitment of the heterochromatin machinery. Tucker et al. consider the functions of the Dhp1 nuclease in termination and gene silencing as separable, with the termination role not an important aspect of its silencing duty. These reports, and that of Kowalik et al.21, have been productive investigations into the role of transcription termination factors in S. pombe gene silencing, a phenomenon that had been explored previously in S. cerevisiae22–24 but which seems mechanistically distinct between the yeast species considering the absence of post-transcriptional gene silencing in S. cerevisiae.\n\n\nPervasive transcription and termination\n\nIn recent years, it has been repeatedly shown that transcription initiation is promiscuous and widespread, with much of the genome capable of being copied into RNA. A corollary to this is the recognition that the termination reaction is an important governor over the transcriptome in its capacity to partition or funnel transcription products into a degradative ‘clean up’ pathway for RNA elimination (such as seen for cryptic unstable transcripts), a maturation pathway for limited processing (such as snoRNAs), or an option to yield spliced and polyadenylated mRNAs.\n\nOne form of pervasive transcription is seen as divergent initiation from promoters. Recent work has gone into characterizing transcripts that radiate in both directions from bidirectionally firing promoters25–29. Non-coding transcripts that extend in the ‘wrong’ (antisense) direction are terminated, and their degradation is tightly coupled to this process30–34. Interestingly, the genomic DNA extending in that direction tends to be enriched in polyA signals and hence termination sequences. In contrast, those in the sense or mRNA-yielding direction are depleted of termination potential because they are enriched in splicing signals, which protect transcription from premature termination35,36. Thus, termination enforces promoter directionality.\n\nA recent genome-wide study showed that prematurely terminated transcripts are one class of RNA that is cleaned up by nuclease surveillance. Mutation of the human nuclear ribonuclease complex, known as the exosome, revealed the surprising frequency with which prematurely terminated RNAs are generated from the genome37, something that might be expected given the imperfect processivity of pol II engaged in transcribing megabase-long genes.\n\nIn an interesting flipping of the conventional way of thinking that termination feeds transcripts to the degradation machinery, a study in S. pombe suggests that the exosome complex can capture a specific paused and backtracked form of pol II and take it to the termination pathway38. Here, the exosome is thought to directly recognize the 3’ end of the extruded nascent transcript in arrested (‘backtracked’) complexes. Yet unresolved is whether the protein sets employed in conventional (polyA-coupled) pol II termination or short non-coding termination (such as the Nrd1-Nab3-Sen1 system in S. cerevisiae) are required for this mode of stopping transcription.\n\nNot only is the regulation of termination important for cellular genomes but also viruses can antagonize normal transcription termination during infection. Using 4-thiouridine labeling of RNA and ribosome profiling of newly made RNA following infection, Rutkowski et al. found that herpes virus inhibits the termination of host, but not viral, transcription39. The resulting readthrough RNAs became observable in a manner similar to those seen in other systems when the exosome or termination machineries become incapacitated by experimental manipulation24,28,37. These aberrant, even polygenic, transcripts were not spliced or translated properly, thus resulting in a form of host shut-off that may favor viral gene expression. How widespread the phenomenon of viral corruption of the integrity of the host’s transcription units is, and its mechanistic basis, remains to be elucidated.\n\nAnother case of a pathologically defeated transcription termination system was observed by Grosso et al. in a human cancer40. Through transcriptome profiling of various renal cell carcinoma samples, they found widespread transcriptional readthrough in many of the samples from patients with some of the worst prognoses. Again, chimeric transcripts spanning transcription units could be identified. The SETD2 gene, which encodes a histone methyltransferase, was frequently inactivated in these cancer lines. Knockdown of SETD2 could recreate the readthrough phenotype. Ectopic expression of the wild-type protein could reverse it, thus implicating chromatin modifications in disease-related failure to terminate effectively. Interestingly, Bcl2, an anti-apoptotic protein often elevated in cancer, was upregulated, suggesting a possible mechanism by which mutation leads to loss of proliferative control.\n\nYet another version of a hybrid readthrough transcript called DoGs (downstream of genes containing transcripts) has been identified41,42. These transcripts have a 5’ end corresponding to an upstream gene and an aberrant 3’ extension of up to tens of kilobases. Hundreds of DoGs were detected following KCl treatment of neuroblastoma cells whereupon their length and abundance increased. These readthrough products, only some of which became polyadenylated, encountered fewer polyA sites downstream of the coding sequence, which could explain less efficient termination for the DoG transcript. It is not clear what functional role the DoGs play; their strong association with chromatin suggests they become incorporated into the nuclear scaffolding during stress responses. Intuitively, this is an appealing model, as osmotic stress shrinks the cell nucleus and collapses chromatin. DoGs may mitigate such deleterious effects, as recently suggested for a set of non-coding, repeat-containing RNAs43. This finding emphasizes that terminator override plays an important role across physiological states and reveals plasticity in our definition of a transcription unit.\n\nThe continued accumulation of instances of pervasive transcription has changed our view of cellular RNA from one that is ‘open reading frame centric’ to one with a greater appreciation for RNA arising from intergenic sequences, either constitutively or under specific physiological or pathological conditions.\n\n\nTranscription termination in immunological systems\n\nThe transcription elongation complex has been of interest to immunologists for many years owing to transcription-dependent mutations introduced into immunoglobulin (Ig)-encoding genes during somatic hypermutation44. Somatic hypermutation is the process by which the variable regions of Ig genes change as the immune response matures, leading to selected antibodies with increasingly higher affinities. A mutagenic enzyme, activation-induced cytidine deaminase (AID), is thought to track with elongating pol II and is associated with premature termination of transcription. This mechanism is proposed to link transcription to a specific hotspot for physiologically important mutations. A recent report extends this model and proposes that pervasive transcripts are functionally important for proper class switch recombination and somatic hypermutation in B lymphocytes45,46. The suggested model is that transcripts in paused, divergently facing elongation complexes become stabilized, offering the complex more of an opportunity to form short DNA-RNA hybrids called R-loops. These loops are part of the transcription termination reaction in which RNA is hybridized to the DNA from which pol II has just departed and can be considered an intermediate on the pathway to disintegration of the transcription bubble and the digestion of the short, anti-sense RNA. B lymphocytes appear to have evolved a mechanism to capture this nucleic acid framework with its exposed single-stranded DNA and make it a target for the AID mutagenesis machinery. In this manner, B cells can focus the genetic changes on the very specific region of Ig proteins that need to vary. We will come back to a discussion of the importance of R-loops in termination in the next section.\n\nWang et al. also present evidence in support of the idea that premature termination provides an opportunity for the single-stranded DNA of the transcription bubble to become a substrate for localized mutation by AID47. They show via chromatin immunoprecipitation that pol II concentrates in the region that undergoes hypermutation. Subsequent manipulation of elongation factor Spt5 by knockdown, which should facilitate premature termination, revealed that DNA strands in the region acquire a single-stranded character and are mutated at higher frequency. Early termination was suggested by the molar abundance of RNA from the 5’ variable region of the heavy chain locus versus the downstream sequences.\n\nIn another immunologically interesting system, the Zfp318 protein has been postulated to regulate termination site selection during the transcription of the gene that encodes both the μ and δ heavy chains in B lymphocytes48. Typically, a precursor transcript is synthesized that encodes the μ heavy chain constant region exons and, further downstream, the corresponding δ constant region exons. Alternative splicing appends one or the other sets of exons to the VDJ-encoding exons, thereby yielding mRNAs for complete IgM or IgD heavy chains. This regulatory event has been known for some time, although the responsible trans-acting factors have been elusive49,50. By engineering a conditional Zfp318 deficiency into the bone marrow lineage of mice, Pioli et al. showed that Zfp318 normally serves to repress an apparent transcriptional termination event at the end of the μ exon series. Thus, there is normally some readthrough into the δ exons, thereby generating a primary transcript that can yield either μ- or δ-encoding mRNA, depending on the splicing pattern48. In the absence of Zfp318, it is suggested that the termination event becomes so strong that transcripts hardly extend into the δ exons, leaving cells with a pool of precursor transcripts ending near the polyA site for the upstream μ exons, and giving B cells no option of making IgD. This was one of the few shifts observed in that transcriptome, showing a very specific consequence following the loss of Zfp318. An independent set of experiments confirmed that incapacitating Zfp318 resulted in a shift from IgD and IgM production to mainly IgM production51. Enders et al. attributed this effect to the rate of pol II elongation and a change in the efficacy of competition between polyadenylation at the end of the μ exons vs. alternative splicing around them to append the δ exons. It will be interesting to learn how the Zfp318 protein, which may bind nucleic acids and could be a key regulator of this process, operates and if it directly influences a true transcription termination event.\n\n\nR-loops, dicer, and genomic stability\n\nThe role of the R-loop in somatic hypermutation and recombination associated with heavy chain class switching is a specialized use of a common feature of the transcription termination zone. One hypothesized function of the senataxin protein, which is the human orthologue of the Sen1 termination factor initially discovered in yeast, is to provoke termination by unraveling the R-loop using a helicase activity1,52. However, these structures may also be involved in coordinating chromatin remodeling, as elucidated in a recent study in which repressive chromatin marks were induced over terminators53. In one example, an RNA duplex is suggested to form from anti-sense transcription of the R-loop, which recruits dicer to the region, leading to trimethylation of H3K9 at the termination zone, thereby linking the termination system to chromatin and the RNAi system in human cells.\n\nInterestingly, Castel et al. show a role for Dicer1 in termination for all three nuclear polymerases, but this function appears to be independent of the rest of the RNAi machinery in S. pombe54. dcr1Δ cells show increased polymerase occupancy at the end of genes, implying stalling of polymerase. Deletion of other RNAi machinery or generating a catalytically dead Dicer1 mutant did not give similar results, suggesting the process is not RNAi mediated. The genes regulated by this mechanism seem to be restricted to those at sites of replicative stress, where high transcription levels result in collision of the transcription bubble and the replication fork. Since DNA-RNA hybrids at these sites are recombinogenic, Dcr1 may be important for maintaining genome integrity, similar to the way senataxin resolves the problem in mammalian cells, although the latter is thought to do so through a helicase activity55,56. The role for dicer in termination could also be related to prior findings that Rnt1, another RNAse III-type enzyme in S. cerevisiae, can provoke termination57.\n\nR-loop-mediated termination is also proposed to play a role in Friedrich’s ataxia. It was suggested that R-loops aberrantly form in the frataxin gene as a result of expansion of GAA repeats in its first intron58,59. Mutated frataxin exhibits features of heterochromatin, H3K9 methylation, and decreased acetylation of H3 and H4, thus a reduced level of expression is to be expected. However, the altered gene also shares many characteristics of canonical R-loop-terminated genes, including a polyA-signal-like sequence upstream of the expansion, followed by a GU-rich sequence similar to the downstream element of polyA signals. This has led to a proposal that the mutated frataxin allele is the victim of premature termination, which contributes to its low level of expression in patients59. While experimental verification is still needed, this is an interesting model for a role of termination in disease.\n\nIn a final case of R-loop involvement in termination, Zhao et al. studied the modification of the repeat region of pol II’s largest subunit60. They found that dimethylation of a specific arginine serves to recruit the survival of motor neuron (SMN) protein to the elongation complex. This protein, in turn, associates with senataxin, which resolves R-loops and recruits the torpedo nuclease Xrn2 during termination. Both SMN and senataxin have been found to suffer mutations in neurodegenerative diseases, again highlighting the importance of the termination reaction in human health and the imperative to understand its molecular basis.\n\nOur understanding of the role of transcription termination across biological systems has expanded significantly with recent advances in experimental techniques and with growing interest in the topic across biological systems. New findings have yielded fresh insight into the longstanding questions of termination mechanism and somatic hypermutation and have provided new paradigms with the identification of cases in which the stringency of termination is relaxed during specific regulatory events or accidentally in pathological examples of dysfunction.\n\n\nAbbreviations\n\nAID, activation-induced cytidine deaminase; DoGs, downstream of genes containing transcripts; Ig, immunoglobulin; lncRNA, long non-coding RNA; mRNA, messenger RNA; pol II, RNA polymerase II; polyA, polyadenylation; SMN, survival of motor neuron protein; snoRNA, small nucleolar RNA.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nWork in the authors’ laboratory was supported by the Emory University School of Medicine and the National Institute of General Medical Sciences.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors thank Dr Graeme Conn for a critical reading of the manuscript and acknowledge insightful comments from the reviewers.\n\n\nReferences\n\nBrow DA: Sen-sing RNA terminators. Mol Cell. 2011; 42(6): 717–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPorrua O, Libri D: Transcription termination and the control of the transcriptome: why, where and how to stop. Nat Rev Mol Cell Biol. 2015; 16(3): 190–202. PubMed Abstract | Publisher Full Text\n\nKuehner JN, Pearson EL, Moore C: Unravelling the means to an end: RNA polymerase II transcription termination. Nat Rev Mol Cell Biol. 2011; 12(5): 283–94. PubMed Abstract | Publisher Full Text\n\nJensen TH, Jacquier A, Libri D: Dealing with pervasive transcription. Mol Cell. 2013; 52(4): 473–84. PubMed Abstract | Publisher Full Text\n\nArndt KM, Reines D: Termination of Transcription of Short Noncoding RNAs by RNA Polymerase II. Annu Rev Biochem. 2015; 84: 381–404. PubMed Abstract | Publisher Full Text\n\nMischo HE, Proudfoot NJ: Disengaging polymerase: terminating RNA polymerase II transcription in budding yeast. Biochim Biophys Acta. 2013; 1829(1): 174–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLogan J, Falck-Pedersen E, Darnell JE Jr, et al.: A poly(A) addition site and a downstream termination region are required for efficient cessation of transcription by RNA polymerase II in the mouse beta maj-globin gene. Proc Natl Acad Sci U S A. 1987; 84(23): 8306–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConnelly S, Manley JL: A functional mRNA polyadenylation signal is required for transcription termination by RNA polymerase II. Genes Dev. 1988; 2(4): 440–52. PubMed Abstract | Publisher Full Text\n\nLibri D: Endless Quarrels at the End of Genes. Mol Cell. 2015; 60(2): 192–4. PubMed Abstract | Publisher Full Text\n\nLuo W, Johnson AW, Bentley DL: The role of Rat1 in coupling mRNA 3'-end processing to transcription termination: implications for a unified allosteric-torpedo model. Genes Dev. 2006; 20(8): 954–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchaughency P, Merran J, Corden JL: Genome-wide mapping of yeast RNA polymerase II termination. PLoS Genet. 2014; 10(10): e1004632. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang H, Rigo F, Martinson HG: Poly(A) Signal-Dependent Transcription Termination Occurs through a Conformational Change Mechanism that Does Not Require Cleavage at the Poly(A) Site. Mol Cell. 2015; 59(3): 437–48. PubMed Abstract | Publisher Full Text\n\nDengl S, Cramer P: Torpedo nuclease Rat1 is insufficient to terminate RNA polymerase II in vitro. J Biol Chem. 2009; 284(32): 21270–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFong N, Brannan K, Erickson B, et al.: Effects of Transcription Elongation Rate and Xrn2 Exonuclease Activity on RNA Polymerase II Termination Suggest Widespread Kinetic Competition. Mol Cell. 2015; 60(2): 256–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSansó M, Levin RS, Lipp JJ, et al.: P-TEFb regulation of transcription termination factor Xrn2 revealed by a chemical genetic screen for Cdk9 substrates. Genes Dev. 2016; 30(1): 117–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlaimo PJ, Shogren-Knaak MA, Shokat KM: Chemical genetic approaches for the elucidation of signaling pathways. Curr Opin Chem Biol. 2001; 5(4): 360–7. PubMed Abstract | Publisher Full Text\n\nShah K, Liu Y, Deirmengian C, et al.: Engineering unnatural nucleotide specificity for Rous sarcoma virus tyrosine kinase to uniquely label its direct substrates. Proc Natl Acad Sci U S A. 1997; 94(8): 3565–70. PubMed Abstract | Free Full Text\n\nJonkers I, Lis JT: Getting up to speed with transcription elongation by RNA polymerase II. Nat Rev Mol Cell Biol. 2015; 16(3): 167–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChalamcharla VR, Folco HD, Dhakshnamoorthy J, et al.: Conserved factor Dhp1/Rat1/Xrn2 triggers premature transcription termination and nucleates heterochromatin to promote gene silencing. Proc Natl Acad Sci U S A. 2015; 112(51): 15548–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTucker JF, Ohle C, Schermann G, et al.: A Novel Epigenetic Silencing Pathway Involving the Highly Conserved 5’-3’ Exoribonuclease Dhp1/Rat1/Xrn2 in Schizosaccharomyces pombe. PLoS Genet. 2016; 12(2): e1005873. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKowalik KM, Shimada Y, Flury V, et al.: The Paf1 complex represses small-RNA-mediated epigenetic gene silencing. Nature. 2015; 520(7546): 248–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVasiljeva L, Kim M, Terzi N, et al.: Transcription termination and RNA degradation contribute to silencing of RNA polymerase II transcription within heterochromatin. Mol Cell. 2008; 29(3): 313–23. PubMed Abstract | Publisher Full Text\n\nHouseley J, Kotovic K, El Hage A, et al.: Trf4 targets ncRNAs from telomeric and rDNA spacer regions and functions in rDNA copy number control. EMBO J. 2007; 26(24): 4996–5006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSteinmetz EJ, Warren CL, Kuehner JN, et al.: Genome-wide distribution of yeast RNA polymerase II and its control by Sen1 helicase. Mol Cell. 2006; 24(5): 735–46. PubMed Abstract | Publisher Full Text\n\nSeila AC, Calabrese JM, Levine SS, et al.: Divergent transcription from active promoters. Science. 2008; 322(5909): 1849–51. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCore LJ, Waterfall JJ, Lis JT: Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science. 2008; 322(5909): 1845–8. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBuratowski S: Transcription. Gene expression--where to start? Science. 2008; 322(5909): 1804–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPreker P, Nielsen J, Kammler S, et al.: RNA exosome depletion reveals transcription upstream of active human promoters. Science. 2008; 322(5909): 1851–4. PubMed Abstract | Publisher Full Text\n\nHe Y, Vogelstein B, Velculescu VE, et al.: The antisense transcriptomes of human cells. Science. 2008; 322(5909): 1855–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlynn RA, Almada AE, Zamudio JR, et al.: Antisense RNA polymerase II divergent transcripts are P-TEFb dependent and substrates for the RNA exosome. Proc Natl Acad Sci U S A. 2011; 108(26): 10460–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrannan K, Kim H, Erickson B, et al.: mRNA decapping factors and the exonuclease Xrn2 function in widespread premature termination of RNA polymerase II transcription. Mol Cell. 2012; 46(3): 311–24. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nReines D: Decapping goes nuclear. Mol Cell. 2012; 46(3): 241–2. PubMed Abstract | Publisher Full Text\n\nNtini E, Järvelin AI, Bornholdt J, et al.: Polyadenylation site-induced decay of upstream transcripts enforces promoter directionality. Nat Struct Mol Biol. 2013; 20(8): 923–8. PubMed Abstract | Publisher Full Text\n\nAlmada AE, Wu X, Kriz AJ, et al.: Promoter directionality is controlled by U1 snRNP and polyadenylation signals. Nature. 2013; 499(7458): 360–3. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nKaida D, Berg MG, Younis I, et al.: U1 snRNP protects pre-mRNAs from premature cleavage and polyadenylation. Nature. 2010; 468(7324): 664–8. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBerg MG, Singh LN, Younis I, et al.: U1 snRNP determines mRNA length and regulates isoform expression. Cell. 2012; 150(1): 53–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSzczepińska T, Kalisiak K, Tomecki R, et al.: DIS3 shapes the RNA polymerase II transcriptome in humans by degrading a variety of unwanted transcripts. Genome Res. 2015; 25(11): 1622–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLemay J, Larochelle M, Marguerat S, et al.: The RNA exosome promotes transcription termination of backtracked RNA polymerase II. Nat Struct Mol Biol. 2014; 21(10): 919–26. PubMed Abstract | Publisher Full Text\n\nRutkowski AJ, Erhard F, L'Hernault A, et al.: Widespread disruption of host transcription termination in HSV-1 infection. Nat Commun. 2015; 6: 7126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrosso AR, Leite AP, Carvalho S, et al.: Pervasive transcription read-through promotes aberrant expression of oncogenes and RNA chimeras in renal carcinoma. eLife. 2015; 4: pii: e09214. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVilborg A, Passarelli MC, Yario TA, et al.: Widespread Inducible Transcription Downstream of Human Genes. Mol Cell. 2015; 59(3): 449–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVilborg A, Steitz JA: Readthrough transcription: How are DoGs made and what do they do? RNA Biol. 2016; 1–5. PubMed Abstract | Publisher Full Text\n\nHall LL, Carone DM, Gomez AV, et al.: Stable C0T-1 repeat RNA is abundant and is associated with euchromatic interphase chromosomes. Cell. 2014; 156(5): 907–19. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nStorb U: Why does somatic hypermutation by AID require transcription of its target genes? Adv Immunol. 2014; 122: 253–77. PubMed Abstract | Publisher Full Text\n\nPefanis E, Wang J, Rothschild G, et al.: Noncoding RNA transcription targets AID to divergently transcribed loci in B cells. Nature. 2014; 514(7522): 389–93. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nPefanis E, Basu U: RNA Exosome Regulates AID DNA Mutator Activity in the B Cell Genome. Adv Immunol. 2015; 127: 257–308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang X, Fan M, Kalis S, et al.: A source of the single-stranded DNA substrate for activation-induced deaminase during somatic hypermutation. Nat Commun. 2014; 5: 4137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPioli PD, Debnath I, Weis JJ, et al.: Zfp318 regulates IgD expression by abrogating transcription termination within the Ighm/Ighd locus. J Immunol. 2014; 193(5): 2546–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoore KW, Rogers J, Hunkapiller T, et al.: Expression of IgD may use both DNA rearrangement and RNA splicing mechanisms. Proc Natl Acad Sci U S A. 1981; 78(3): 1800–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaki R, Roeder W, Traunecker A, et al.: The role of DNA rearrangement and alternative RNA processing in the expression of immunoglobulin delta genes. Cell. 1981; 24(2): 353–65. PubMed Abstract | Publisher Full Text\n\nEnders A, Short A, Miosge LA, et al.: Zinc-finger protein ZFP318 is essential for expression of IgD, the alternatively spliced Igh product made by mature B lymphocytes. Proc Natl Acad Sci U S A. 2014; 111(12): 4513–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSkourti-Stathaki K, Proudfoot NJ, Gromak N: Human senataxin resolves RNA/DNA hybrids formed at transcriptional pause sites to promote Xrn2-dependent termination. Mol Cell. 2011; 42(6): 794–805. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nSkourti-Stathaki K, Kamieniarz-Gdula K, Proudfoot NJ: R-loops induce repressive chromatin marks over mammalian gene terminators. Nature. 2014; 516(7531): 436–9. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nCastel SE, Ren J, Bhattacharjee S, et al.: Dicer promotes transcription termination at sites of replication stress to maintain genome stability. Cell. 2014; 159(3): 572–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuraweera A, Becherel OJ, Chen P, et al.: Senataxin, defective in ataxia oculomotor apraxia type 2, is involved in the defense against oxidative DNA damage. J Cell Biol. 2007; 177(6): 969–79. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHatchi E, Skourti-Stathaki K, Ventz S, et al.: BRCA1 recruitment to transcriptional pause sites is required for R-loop-driven DNA damage repair. Mol Cell. 2015; 57(4): 636–47. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nGhazal G, Gagnon J, Jacques P, et al.: Yeast RNase III triggers polyadenylation-independent transcription termination. Mol Cell. 2009; 36(1): 99–109. PubMed Abstract | Publisher Full Text\n\nGroh M, Lufino MM, Wade-Martins R, et al.: R-loops associated with triplet repeat expansions promote gene silencing in Friedreich ataxia and fragile X syndrome. PLoS Genet. 2014; 10(5): e1004318. PubMed Abstract | Publisher Full Text | Free Full Text\n\nButler JS, Napierala M: Friedreich's ataxia--a case of aberrant transcription termination? Transcription. 2015; 6(2): 33–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao DY, Gish G, Braunschweig U, et al.: SMN and symmetric arginine dimethylation of RNA polymerase II C-terminal domain control termination. Nature. 2016; 529(7584): 48–53. PubMed Abstract | Publisher Full Text | F1000 Recommendation"
}
|
[
{
"id": "14553",
"date": "23 Jun 2016",
"name": "David Brow",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14554",
"date": "23 Jun 2016",
"name": "Dong Wang",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1478
|
https://f1000research.com/articles/3-238/v1
|
08 Oct 14
|
{
"type": "Review",
"title": "Back to the drawing board: Re-thinking the role of GLI1 in pancreatic carcinogenesis",
"authors": [
"Tara L. Hogenson",
"Matthias Lauth",
"Marina Pasca diMagliano",
"Martin E. Fernandez-Zapico",
"Tara L. Hogenson",
"Matthias Lauth",
"Marina Pasca diMagliano"
],
"abstract": "Aberrant activation of the transcription factor GLI1, a central effector of the Hedgehog (HH) pathway, is associated with several malignancies, including pancreatic ductal adenocarcinoma (PDAC), one of most deadly human cancers. GLI1 has been described as an oncogene in PDAC, making it a promising target for drug therapy. Surprisingly, clinical trials targeting HH/GLI1 axis in advanced PDAC were unsuccessful, leaving investigators questioning the mechanism behind these failures. Recent evidence suggests the loss of GLI1 in the later stages of PDAC may actually accelerate disease. This indicates GLI1 may play a dual role in PDAC, acting as an oncogene in the early stages of disease and a tumor-suppressor in the late stages.",
"keywords": [
"The protein GLI1",
"originally isolated in 1987 due to high levels of amplification in malignant glioma (Kinzler et al.",
"1987)",
"is a member of the GLI family of transcription factors. This family also includes GLI2 and GLI3. GLI1 is highly conserved from Drosophila to humans and is required for developmental response via transcriptional regulation of target genes (Dennler et al.",
"2007",
"Hui & Angers",
"2011",
"Javelaud et al.",
"2012). The GLI proteins",
"including GLI1",
"are transcriptional mediators of Hedgehog (HH) signaling",
"and regulate multiple cellular processes such as cell fate determination",
"tissue patterning",
"proliferation and transformation",
"which give this transcription factor a significant role in carcinogenesis if deregulated (Dennler et al.",
"2007",
"Hui & Angers",
"2011",
"Javelaud et al.",
"2012). GLI1 is expressed in different human malignancies including pancreatic ductal adenocarcinoma (PDAC) (Eberl et al.",
"2012",
"Fiaschi et al.",
"2009",
"Goel et al.",
"2013",
"Hui & Angers",
"2011",
"Mills et al.",
"2013",
"Rajurkar et al.",
"2012",
"Thayer et al.",
"2003). In PDAC",
"GLI1 is prevalently expressed in the stroma",
"in response to HH ligands secreted by the epithelial cells (Yauch et al.",
"2008). However",
"lower epithelial expression of Gli1 has also been reported",
"possibly with non-canonical functions (Nolan-Stevaux et al.",
"2009)."
],
"content": "Introduction\n\nThe protein GLI1, originally isolated in 1987 due to high levels of amplification in malignant glioma (Kinzler et al., 1987), is a member of the GLI family of transcription factors. This family also includes GLI2 and GLI3. GLI1 is highly conserved from Drosophila to humans and is required for developmental response via transcriptional regulation of target genes (Dennler et al., 2007; Hui & Angers, 2011; Javelaud et al., 2012). The GLI proteins, including GLI1, are transcriptional mediators of Hedgehog (HH) signaling, and regulate multiple cellular processes such as cell fate determination, tissue patterning, proliferation and transformation, which give this transcription factor a significant role in carcinogenesis if deregulated (Dennler et al., 2007; Hui & Angers, 2011; Javelaud et al., 2012). GLI1 is expressed in different human malignancies including pancreatic ductal adenocarcinoma (PDAC) (Eberl et al., 2012; Fiaschi et al., 2009; Goel et al., 2013; Hui & Angers, 2011; Mills et al., 2013; Rajurkar et al., 2012; Thayer et al., 2003). In PDAC, GLI1 is prevalently expressed in the stroma, in response to HH ligands secreted by the epithelial cells (Yauch et al., 2008). However, lower epithelial expression of Gli1 has also been reported, possibly with non-canonical functions (Nolan-Stevaux et al., 2009).\n\n\nGLI1 as an oncogene in PDAC\n\nGLI1 plays a key role in PDAC initiation by modulating the activity of two different cellular compartments, the epithelium and stroma. Rajurkar et al. demonstrated that targeted overexpression of GLI1 in the pancreas epithelium accelerates PDAC initiation by KRAS, a small GTPase mutated in more than 90% of PDAC cases (Rajurkar et al., 2012). Through use of a mouse model with simultaneous activation of oncogenic KRAS and inhibition of GLI1 in the pancreas epithelium, this group also demonstrated that decreased GLI1 activity reduced the incidence of KRAS-driven PDAC precursor lesions (pancreatic intraepithelial neoplasias or PanINs) and PDAC. Similarly, Mills and colleagues using a mouse model for pancreas-specific oncogenic KRAS expression (KC mice) bred on a Gli1 null background (GKO/KC) defined a key role for GLI1 on PDAC initiation through the modulation of the activity of fibroblasts (Mills et al., 2013). The KC mice developed PanIN lesions with 100% penetrance and PDAC in advanced age, while the GKO/KC mice did not develop PDAC and had increased survival rate when compared to KC mice. Histopathological analysis of the pancreata showed KC mice developed PanIN lesions and PDAC, while 80% of GKO/KC had normal pancreata.\n\nAnalysis of the molecular mechanism underlying this phenomenon reveals that GLI1 both regulates different target genes and is modulated by different signaling pathways depending on the cellular compartment. For instance, GLI1 activity is mainly modulated by the canonical HH signaling in fibroblasts (Yauch et al., 2008). This cascade is activated by binding of the ligand to the receptor Patched (Ptch), resulting in activation of the G-coupled receptor, Smoothened (Smo) (Javelaud et al., 2012; Yauch et al., 2008). Once activated, Smo induces GLI1 activation and upregulation of its target genes (Hui & Angers, 2011; McMahon et al., 2003). The HH ligand, Sonic Hedgehog (SHH), and components of the HH signaling pathway, including Ptch and Smo, are undetectable in the normal pancreas but overexpressed in PanINs and PDAC (Thayer et al., 2003). Inhibition of the HH pathway in PDAC cell-based xenograft models through Smo inhibition has been shown to reduce GLI1 activity and tumor growth (Feldmann et al., 2007; Thayer et al., 2003; Yauch et al., 2008). In addition, genomic sequencing of human pancreatic cancer samples revealed widespread mutations consistent with activation of the Hedgehog signaling pathway (Jones et al., 2008). While the association between HH activity and pancreatic cancer has been described over a decade ago, there is still uncertainty as to the downstream effect of HH activation in this disease. Mills et al. identified the cytokine IL-6 as a HH/GLI1 target gene in pancreatic fibroblasts (Mills et al., 2013). Increased IL-6 expression in the stromal compartment induces activation of STAT3 in the neighboring cancer cells, an essential molecular event for the progression of premalignant lesions in PDAC (Figure 1).\n\nDuring the early stages of PDAC, GLI1 is activated in the fibroblasts through canonical HH signaling. GLI1 promotes expression of the cytokine IL-6, which stimulates expression of STAT3 in neighboring cancer cells, promoting the progression of PanIN lesions to PDAC. In the later stages of PDAC, GLI1 binds the FASL promoter and regulates the expression of this ligand in the fibroblast, leading to higher levels of apoptosis in these tumors. In addition, in cancer cells, GLI1 induces the expression of FAS and CDH1 expression, leading to a tumor protective effect.\n\n\nHedgehog-independent mechanisms for GLI1 expression in PDAC\n\nWhile dysregulation of HH-GLI1 signaling has been shown to play an important role in PDAC formation, several studies have demonstrated that GLI1 expression can be activated through HH-independent mechanisms in PDAC, particularly in the epithelial compartment (Dennler et al., 2007; Eberl et al., 2012; Goel et al., 2013; Ji et al., 2007; Nolan-Stevaux et al., 2009; Nye et al., 2014). Nolan-Stevaux et al. demonstrated that deletion of Smo receptor in pancreatic epithelium had no effect on KRAS induced tumor formation, nor on GLI1 expression in epithelial cells (Nolan-Stevaux et al., 2009). This indicates a Smo-independent mechanism for GLI1 regulation in PDAC cells downstream of KRAS. In fact, Ji et al. demonstrated that KRAS is a modulator of GLI1 activity and requires the transcription factor for PDAC growth in vitro (Ji et al., 2007). In the epithelial compartment, GLI1 is regulated in a HH-independent manner, downstream of KRAS. Accordingly, Ji et al. showed that Gli1 protein degradation is blocked in a MAPK-dependent manner. Furthermore, Rajurkar et al. showed a role for GLI1 in the regulation of the NF-κB pathway, a signaling cascade linked to PDAC development (Algul et al., 2002; Ougolkov et al., 2005; Pan et al., 2008; Wang et al., 1999), downstream of KRAS (Rajurkar et al., 2012). This group has identified the I-kappa-B kinase epsilon (IKBKE)/NF-κB pathway as a direct target of the GLI1 mediating KRAS-dependent pancreatic epithelial transformation in vivo (Junhao Mao, University of Massachusetts and Martin E. Fernandez-Zapico personal communication).\n\nPDAC is characterized by a dense desmoplastic reaction associated with the primary tumor. The abundance of connective tissue is due to an increase in growth factor production in the tumor microenvironment through autocrine and paracrine oncogenic signaling pathways (Mahadevan & Von Hoff, 2007). Oncogenic KRAS activates SHH production, but HH ligands do not activate the HH pathway in tumor epithelial cells in an autocrine manner (Lauth et al., 2010; Mills et al., 2013; Yauch et al., 2008). HH signaling in PDAC occurs in a paracrine fashion where HH signaling from PDAC cells to stromal cells has been shown to promote desmoplasia (Yauch et al., 2008). Lauth et al. demonstrated that this shift from autocrine to paracrine signaling is through activation of the RAS effector dual specificity tyrosine phosphorylated and regulated kinase 1B (DYRK1B) (Lauth et al., 2010). The authors proposed this is achieved through DYRK1B inhibition of GLI2 function and promotion of the repressor GLI3, and subsequent inhibition of GLI1, in PDAC cells.\n\nTGFβ has been shown to promote GLI1 expression in pancreatic cancer cells (Nolan-Stevaux et al., 2009). TGFβ induces the expression of GLI1 through Smad3 and LET-dependent upregulation of GLI2 independent of HH signaling (Dennler et al., 2007; Dennler et al., 2009). Nye et al. demonstrated that TGFβ, in addition to controlling GLI1 expression, can also modulate its activity by promoting the formation of a transcriptional complex with the TGFβ-regulated transcription factors, SMAD2 and 4, and the histone acetyltransferase, PCAF, in cancer cells to regulate TGFβ-induced gene expression (Nye et al., 2014). TGFβ induced GLI2 expression, and subsequent GLI1 activation, is associated with epithelial to mesenchymal transition (EMT), tumor growth, and metastasis (Javelaud et al., 2012).\n\nIn addition to TGFβ and KRAS activation, epidermal growth factor receptor (EGFR) signaling, a cascade aberrantly activated in the majority of PDACs, has been demonstrated to play a critical role in HH/GLI1-regulated tumor-initiating pancreatic cancer cells (Eberl et al., 2012). Eberl and colleagues demonstrated EGFR and HH act together to promote cancer cell proliferation by modulating gene expression through a GLI1-dependent mechanism. This suggests HH/GLI1 signaling works synergistically through distinct novel pathways during tumor initiation and growth.\n\n\nClinical trials targeting the hedgehog/GLI1 axis in PDAC\n\nThe concept that HH/GLI1 signaling might be required for PDAC growth, hence a suitable therapeutic target, has been first validated in a genetically engineered mouse model of pancreatic cancer that combines expression of oncogenic Kras with mutation of the tumor suppressor p53, the KPC mouse (Hingorani et al., 2005). Treatment of KPC mice with a Smo inhibitor in combination with gemcitabine led to a moderate but significant increase in survival (Feldmann et al., 2008; Olive et al., 2009). A preclinical study of the HH inhibitor, saridegib (IPI-926), co-administered with gemcitabine, produced a transient increase in vascular density, increased chemotherapy drug delivery, and improved disease stabilization in pancreatic cancer cells (Olive et al., 2009). Based on these results, phase II clinical trials were approved evaluating saridegib and an additional Hh inhibitor, vismodegib (GDC-0449), for treatment of pancreatic cancer. Surprisingly, the clinical trial for both vismodegib and saridegib showed a higher rate of progressive disease when compared to placebo (Catenacci et al., 2013). Similar findings were seen in a separate phase I trial of vismodegib in 8 patients with pancreatic cancer (LoRusso et al., 2011). Although hedgehog inhibitors have been successful for treating basal cell carcinoma and medulloblastoma, they do not appear to have the same effect in advanced pancreatic cancer.\n\nThese disappointing results left investigators questioning the molecular mechanism responsible for these failed clinical trials. Although there is overwhelming evidence that GLI1 plays an important role in tumor initiation and progression of several kinds of malignancies, these results suggest the transcription factor may have a tumor protective role in the later stages of certain cancers. In fact, recent studies investigating GLI1 expression in PDAC have revealed GLI1 may switch from a tumor promoting to a tumor protective molecule in the later stages of PDAC.\n\n\nGLI1 as a tumor suppressor in PDAC\n\nIn contrast to the current paradigm for GLI1 expression and tumor progression, one study found GLI1 expression may actually decrease cell motility in advanced PDAC (Joost et al., 2012). Joost et al. demonstrated that GLI1 regulates epithelial differentiation through transcriptional activation of the cell adhesion molecule, E-Cadherin (CDH1), in PDAC cells. Lowered expression of GLI1 in PDAC cells lead to a loss of CDH1 expression and promotion of EMT. The transition from epithelial to motile mesenchymal cells is thought to be a critical event for metastasis of carcinomas. Decreased expression of CDH1 is associated with increased metastasis and invasion, while increased expression is associated with lower tumor malignancy (Seidel et al., 2004; von Burstin et al., 2009). PDAC is strongly associated with early invasion and metastasis. Loss of GLI1 was also shown to decrease expression of additional important epithelial marker genes, including Keratin 19 (KRT19) and adherens junctions components EVA1 and PTPRM, leading to increased cell motility. This indicates that as PDAC progresses, lower GLI1 levels may actually prime tumor cells towards an EMT program, which would be associated with metastasis and advanced stages of the disease.\n\nMills et al. examined the role of GLI1 expression in the later stages of PDAC using a mouse model for advanced pancreatic cancer (Mills et al., 2014). In this study, the loss of GLI1 actually accelerated PDAC progression during the later stages of tumorigenesis. PDAC mice lacking GLI1 showed reduced survival when compared to GLI1 wild type littermates. While both cohorts of mice displayed the common features of advanced PDAC, loss of GLI1 was associated with decreased survival and increased tumor burden. Analysis of the mechanism revealed the pro-apoptotic FAS/FASL axis as a potential mediator for this phenomenon. Loss of GLI1 was associated with a significant decrease in expression of FAS/FASL, leading to lower apoptosis levels and increased tumor progression (Figure 1).\n\nIn agreement with these findings, two recent studies demonstrated that the deletion of the GLI1 inducer SHH, using a mouse model for PDAC, led to more aggressive tumors (Lee et al., 2014; Rhim et al., 2014). Interestingly, Rhim’s study reported the occurrence of poorly differentiated tumors, with increased vascularity, and significantly reduced stromal content. In contrast, the Lee paper only described a modest reduction in the stromal compartment. The current paradigm for PDAC is that the tumor stroma plays an important role in promotion of neoplastic growth and progression since PDAC is typically associated with a dense desmoplastic reaction. However, the Rhim study shows that tumors with reduced stroma may display a more aggressive behavior than those with an extensive stromal compartment. This concept is further supported by a recent report demonstrating that tumor stroma restrains pancreatic cancer progression and that pharmacological HH pathway activation in stromal cells can actually slow down in vivo tumorigenesis (Lee et al., 2014). The complexity of these findings reflects our incomplete understanding of the precise biological role of HH/GLI1 signaling in pancreatic cancer. In fact, the level of activation of HH signaling might induce different biological responses during the carcinogenesis process, as commonly observed during embryonic development. In fact, manipulation of the membrane mediators of HH signaling to reduce HH signaling leads to an increase of angiogenesis with low HH levels, but not with complete inhibition. Intriguingly, the Rhim and Lee studies generated a low HH signaling environment by eliminating SHH, but not IHH, another HH ligand expressed in pancreatic cancer. Similarly, studies altering the expression of Gli1 leave intact the other mediators of HH signaling, GLI2 and GLI3 (the latter mainly an inhibitor of HH target genes). It remains to be seen if manipulating GLI1 levels within the epithelial tumor compartment in later stages of disease is of any therapeutic value. Based on work from Fendrich et al. on HH signaling and acinar cell differentiation, it might even be provocatively proclaimed that increasing GLI1 levels could drive terminal differentiation and thus result in lower tumorigenicity (Fendrich et al., 2008).\n\n\nDiscussion\n\nThese studies demonstrating GLI1 may act as a tumor suppressor in the late stage of PDAC give insight into the disappointing results of clinical trials testing HH inhibitors in metastatic PDAC patients. While the Olive experiments reported acute administration of IPI-926 increased survival due to decreased stromal content and increased vascularity, the HH inhibitor performed poorly in pancreatic cancer clinical trials in patients. One explanation for this discrepancy may be the short duration of treatment (3 weeks) in the Olive’s experiments, which may have not accurately detected disease progression following HH inhibition. This indicates that as PDAC progresses, the initial positive effects of HH inhibition may be eliminated as GLI1 levels decrease. In Rhim’s study, the authors discovered that SHH and GLI1 deficient tumors were more aggressive, poorly differentiated, and exhibited increased vascularity (Rhim et al., 2014). This suggests HH/GLI1 pathway inhibition may have a proangiogenic effect. Due to the increase in vascularity of the SHH deficient mice tumors, the authors investigated the effect of angiogenesis inhibition by administering anti-VEGF to tumor-bearing SHH deficient mice. This therapy led to a significant improvement in the overall survival of mice with undifferentiated tumors. Based on this response, the subset of PDAC patients with undifferentiated tumors may benefit from anti-angiogenic therapy. In summary, due to the high complexity of PDAC initiation and progression, a personalized strategy for treatment should be considered. Under this strategy, PDAC should be analyzed before treatment to determine expression of GLI1 and upstream regulators in order to better define therapeutic options.",
"appendix": "Author contributions\n\n\n\nT.L.H and M.E.F.-Z. developed the concept of the review and T.L.H, M.L., M.P.M. and M.E.F.-Z. wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by National Institutes of Health Grant CA136526, Division of Oncology Research (Mayo Clinic), Mayo Clinic Pancreatic SPORE P50 Grant CA102701 and Mayo Clinic Center for Cell Signaling in Gastroenterology Grant P30 DK84567 (to M.E.F.-Z.).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank Emily Porcher and Pam Becker for secretarial assistance.\n\n\nReferences\n\nAlgul H, Adler G, Schmid RM: NF-kappaB/Rel transcriptional pathway: implications in pancreatic cancer. Int J Gastrointest Cancer. 2002; 31(1–3): 71–78. PubMed Abstract | Publisher Full Text\n\nCatenacci D, Bahary N, Sreenivasa RN, et al.: Final analysis of a phase IB/randomized phase II study of gemcitabine (G) plus placebo (P) or vismodegib (V), a hedgehog (Hh) pathway inhibitor, in patients (pts) with metastatic pancreatic cancer (PC): A University of Chicago phase II consortium study. J Clin Oncol. 2013; 31: (suppl; abstr 4012). Reference Source\n\nDennler S, Andre J, Alexaki I, et al.: Induction of sonic hedgehog mediators by transforming growth factor-beta: Smad3-dependent activation of Gli2 and Gli1 expression in vitro and in vivo. Cancer Res. 2007; 67(14): 6981–6986. PubMed Abstract | Publisher Full Text\n\nDennler S, Andre J, Verrecchia F, et al.: Cloning of the human GLI2 Promoter: transcriptional activation by transforming growth factor-beta via SMAD3/beta-catenin cooperation. J Biol Chem. 2009; 284(46): 31523–31531. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEberl M, Klingler S, Mangelberger D, et al.: Hedgehog-EGFR cooperation response genes determine the oncogenic phenotype of basal cell carcinoma and tumour-initiating pancreatic cancer cells. EMBO Mol Med. 2012; 4(3): 218–233. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeldmann G, Dhara S, Fendrich V, et al.: Blockade of hedgehog signaling inhibits pancreatic cancer invasion and metastases: a new paradigm for combination therapy in solid cancers. Cancer Res. 2007; 67(5): 2187–2196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeldmann G, Habbe N, Dhara S, et al.: Hedgehog inhibition prolongs survival in a genetically engineered mouse model of pancreatic cancer. Gut. 2008; 57(10): 1420–1430. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFendrich V, Esni F, Garay MV, et al.: Hedgehog signaling is required for effective regeneration of exocrine pancreas. Gastroenterology. 2008; 135(2): 621–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFiaschi M, Rozell B, Bergström A, et al.: Development of mammary tumors by conditional expression of GLI1. Cancer Res. 2009; 69(11): 4810–4817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoel HL, Pursell B, Chang C, et al.: GLI1 regulates a novel neuropilin-2/α6β1 integrin based autocrine pathway that contributes to breast cancer initiation. EMBO Mol Med. 2013; 5(4): 488–508. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHingorani SR, Wang L, Multani AS, et al.: Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell. 2005; 7(5): 469–83. PubMed Abstract | Publisher Full Text\n\nHui CC, Angers S: Gli proteins in development and disease. Annu Rev Cell Dev Biol. 2011; 27: 513–537. PubMed Abstract | Publisher Full Text\n\nJavelaud D, Pierrat MJ, Mauviel A: Crosstalk between TGF-β and hedgehog signaling in cancer. FEBS Lett. 2012; 586(14): 2016–2025. PubMed Abstract | Publisher Full Text\n\nJi Z, Mei FC, Xie J, et al.: Oncogenic KRAS activates hedgehog signaling pathway in pancreatic cancer cells. J Biol Chem. 2007; 282(19): 14048–14055. PubMed Abstract | Publisher Full Text\n\nJones S, Zhang X, Parsons DW, et al.: Core signaling pathways in human pancreatic cancers revealed by global genomic analyses. Science. 2008; 321(5897): 1801–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoost S, Almada LL, Rohnalter V, et al.: GLI1 inhibition promotes epithelial-to-mesenchymal transition in pancreatic cancer cells. Cancer Res. 2012; 72(1): 88–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKinzler KW, Bigner SH, Bigner DD, et al.: Identification of an amplified, highly expressed gene in a human glioma. Science. 1987; 236(4797): 70–73. PubMed Abstract | Publisher Full Text\n\nLauth M, Bergstrom A, Shimokawa T, et al.: DYRK1B-dependent autocrine-to-paracrine shift of Hedgehog signaling by mutant RAS. Nat Struct Mol Biol. 2010; 17(6): 718–725. PubMed Abstract | Publisher Full Text\n\nLee JJ, Perera RM, Wang H, et al.: Stromal response to Hedgehog signaling restrains pancreatic cancer progression. Proc Natl Acad Sci U S A. 2014; 111(30): E3091–3100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoRusso PM, Rudin CM, Reddy JC, et al.: Phase I trial of hedgehog pathway inhibitor vismodegib (GDC-0449) in patients with refractory, locally advanced or metastatic solid tumors. Clin Cancer Res. 2011; 17(8): 2502–2511. PubMed Abstract | Publisher Full Text\n\nMahadevan D, Von Hoff DD: Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol Cancer Ther. 2007; 6(4): 1186–1197. PubMed Abstract | Publisher Full Text\n\nMcMahon AP, Ingham PW, Tabin CJ, et al.: Developmental roles and clinical significance of hedgehog signaling. Curr Top Dev Biol. 2003; 53: 1–114. PubMed Abstract | Publisher Full Text\n\nMills L, Zhang DL, Marler R, et al.: Inactivation of the transcription factor GLI1 accelerates pancreatic cancer progression. J Biol Chem. 2014; 289(23): 16516–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMills LD, Zhang Y, Marler RJ, et al.: Loss of the transcription factor GLI1 identifies a signaling network in the tumor microenvironment mediating KRAS oncogene-induced transformation. J Biol Chem. 2013; 288(17): 11786–11794. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNolan-Stevaux O, Lau J, Truitt ML, et al.: GLI1 is regulated through Smoothened-independent mechanisms in neoplastic pancreatic ducts and mediates PDAC cell survival and transformation. Genes Dev. 2009; 23(1): 24–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNye MD, Almada LL, Fernandez-Barrena MG, et al.: The transcription factor GLI1 interacts with SMAD proteins to modulate transforming growth factor β-induced gene expression in a p300/CREB-binding protein-associated factor (PCAF)-dependent manner. J Biol Chem. 2014; 289(22): 15495–506. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOlive KP, Jacobetz MA, Davidson CJ, et al.: Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science. 2009; 324(5933): 1457–1461. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOugolkov AV, Fernandez-Zapico ME, Savoy DN, et al.: Glycogen synthase kinase-3beta participates in nuclear factor kappaB-mediated gene transcription and cell survival in pancreatic cancer cells. Cancer Res. 2005; 65(6): 2076–2081. PubMed Abstract | Publisher Full Text\n\nPan X, Arumugam T, Yamamoto T, et al.: Nuclear factor-kappaB p65/relA silencing induces apoptosis and increases gemcitabine effectiveness in a subset of pancreatic cancer cells. Clin Cancer Res. 2008; 14(24): 8143–8151. PubMed Abstract | Publisher Full Text\n\nRajurkar M, De Jesus-Monge WE, Driscoll DR, et al.: The activity of Gli transcription factors is essential for Kras-induced pancreatic tumorigenesis. Proc Natl Acad Sci U S A. 2012; 109(17): E1038–1047. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRhim AD, Oberstein PE, Thomas DH, et al.: Stromal elements act to restrain, rather than support, pancreatic ductal adenocarcinoma. Cancer Cell. 2014; 25(6): 735–747. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSeidel B, Braeg S, Adler G, et al.: E- and N-cadherin differ with respect to their associated p120ctn isoforms and their ability to suppress invasive growth in pancreatic cancer cells. Oncogene. 2004; 23(32): 5532–5542. PubMed Abstract | Publisher Full Text\n\nThayer SP, di Magliano MP, Heiser PW, et al.: Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature. 2003; 425(6960): 851–856. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvon Burstin J, Eser S, Paul MC, et al.: E-cadherin regulates metastasis of pancreatic cancer in vivo and is suppressed by a SNAIL/HDAC1/HDAC2 repressor complex. Gastroenterology. 2009; 137(1): 361–371, 371.e1–5. PubMed Abstract | Publisher Full Text\n\nWang W, Abbruzzese JL, Evans DB, et al.: The nuclear factor-kappa B RelA transcription factor is constitutively activated in human pancreatic adenocarcinoma cells. Clin Cancer Res. 1999; 5(1): 119–127. PubMed Abstract\n\nYauch RL, Gould SE, Scales SJ, et al.: A paracrine requirement for hedgehog signalling in cancer. Nature. 2008; 455(7211): 406–410. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "6370",
"date": "16 Oct 2014",
"name": "Alain Mauviel",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article Hogenson et al. provide us with a timely review regarding the current knowledge about the role of the transcription factor GLI1 in pancreatic carcinoma. There is an important focus on the dual activity of GLI1 during pancreatic carcinogenesis depending on the stage of disease progression, as evidenced in most recent works in this field, both clinical and experimental, that are nicely summarized in this review.Another important aspect of this review consists in integrating GLI1 as a transcription factor that is not solely regulated by Hedgehog signaling downstream of SMO but also by other pro-tumorigenic pathways, such as TGF-beta and KRAS signaling.Overall, I believe that this manuscript is very focused and contains valuable information for the broader readership, with up-to-date citation of the most relevant and recent literature in the field.A couple of minor points may be corrected or improved:A figure summarizing the role of GLI1 downstream of the various pathways described to modulate its expression/activity would be helpful. On page 3, one reads \"TGFβ induces the expression of GLI1 through Smad3 and LET-dependent up regulation of GLI2\". \"LET-dependent\" should be replaced by beta-catenin/LEF-TCF-dependent or something similar.",
"responses": [
{
"c_id": "2028",
"date": "23 Jun 2016",
"name": "Martin Fernandez-Zapico",
"role": "Author Response",
"response": "A figure summarizing the role of GLI1 downstream of the various pathways described to modulate its expression/activity would be helpful. Response: An additional figure was added to the review (Figure 1) to describe the various pathways that modulate GLI1 expression as discussed in the review. Figure 1 from the first version of this article was changed to Figure 2. On page 3, one reads \"TGFβ induces the expression of GLI1 through Smad3 and LET-dependent up regulation of GLI2\". \"LET-dependent\" should be replaced by beta-catenin/LEF-TCF-dependent or something similar. Response: “LET-dependent” was corrected to “β-catenin/LEF-TCF-dependent” on page 3."
}
]
},
{
"id": "6367",
"date": "21 Oct 2014",
"name": "Natalia Riobo",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe review is timely and addresses a very important problem in pancreatic cancer. The Smoothened inhibitors that work well for other tumor types have not only failed to stop the progression, but instead promote aggressive behavior in pancreatic cancer. The authors make a good case of balancing the evidence that suggests that there is HH-independent upregulation of GLI1 in the epithelial cells and a Hh-dependent upregulation in fibroblasts. Moreover, they nicely discuss how GLI1 is necessary for PanIN formation and then restrains further cancer progression. What is lacking in the review is a consideration of potential non-canonical effects of the Smo inhibitors. It is known that Smo induces cytoskeletal changes in fibroblasts, for instance, and that it can regulate glucose uptake in other cell types. Perhaps modulation of GLI1 is a bystander effect confusing the results. And the GLI1 knockout animals only partly acknowledge this interpretation, since some effects can be cell-type specific and opposing, as discussed in the review. A minor criticism is the following: the introduction erroneously says that GLI1 is conserved from Drosophila to humans. However, GLI3 is the closest homolog in sequence and function to Drosophila Ci. It seems that the authors meant the GLI family is conserved.",
"responses": [
{
"c_id": "2027",
"date": "23 Jun 2016",
"name": "Martin Fernandez-Zapico",
"role": "Author Response",
"response": "What is lacking in the review is a consideration of potential non-canonical effects of the Smo inhibitors. It is known that Smo induces cytoskeletal changes in fibroblasts, for instance, and that it can regulate glucose uptake in other cell types. Perhaps modulation of GLI1 is a bystander effect confusing the results. And the GLI1 knockout animals only partly acknowledge this interpretation, since some effects can be cell-type specific and opposing, as discussed in the review. Response: Additional considerations were added concerning the use of SMO inhibitors and downstream GLI-independent effects to the Discussion section. A minor criticism is the following: the introduction erroneously says that GLI1 is conserved from Drosophila to humans. However, GLI3 is the closest homolog in sequence and function to Drosophila Ci. It seems that the authors meant the GLI family is conserved. Response: This statement was corrected to “the GLI family of transcription factors is highly conserved and is required for developmental response via transcriptional regulation of target genes” in the Introduction. The statement regarding Drosophila to humans was removed."
}
]
},
{
"id": "6366",
"date": "23 Oct 2014",
"name": "Barbara Stecca",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis review provides a comprehensive summary about the role of the transcription factor GLI1 in pancreatic cancer (PDAC). This manuscript highlights the dual role of GLI1 during pancreatic carcinogenesis, acting as an oncogene in the early stages of disease and as a tumor-suppressor in the late stages. Recent evidence suggests the loss of GLI1 in the later stages of PDAC might accelerate disease progression. This might explain why Smoothened (SMO) inhibitors have been successful for treating basal cell carcinoma and medulloblastoma, but do not appear to have the same effect in metastatic PDAC. Moreover, this article summarizes recent data on the integration of GLI1 with other signaling pathways, suggesting that GLI1 is not only regulated by the upstream Hedgehog signaling in a SMO-dependent manner, but also by other oncogenic inputs, such as KRAS, TGF-beta and EGFR signaling. Recent experimental data suggest that lower GLI1 levels associate with PDAC progression, whereas increasing GLI1 levels could drive terminal differentiation and decreased PDAC tumorigenicity. What is missing in this review is a consideration about the factors that might contribute to decrease GLI1 levels and activity during PDAC progression. I would also suggest to mention in the Discussion that the dual role of GLI1 is so far limited to PDAC, as the evidence is lacking in other cancer types.",
"responses": [
{
"c_id": "2026",
"date": "23 Jun 2016",
"name": "Martin Fernandez-Zapico",
"role": "Author Response",
"response": "Recent experimental data suggest that lower GLI1 levels associate with PDAC progression, whereas increasing GLI1 levels could drive terminal differentiation and decreased PDAC tumorigenicity. What is missing in this review is a consideration about the factors that might contribute to decrease GLI1 levels and activity during PDAC progression. Response: We added additional considerations regarding other factors that may drive down GLI1 levels during PDAC progression to the “Discussion” section. Mechanisms included are activation of DYRK1B kinase, which promotes a shift from autocrine to paracrine signaling, which may lead to decreased GLI1 expression. Also, decreased HH signaling in advanced PDAC may allow for increased expression of GLI1 repressors. I would also suggest to mention in the Discussion that the dual role of GLI1 is so far limited to PDAC, as the evidence is lacking in other cancer types. Response: The following statement was added to the final paragraph of the Discussion section “Further studies are needed to fully understand the role of GLI1 in PDAC carcinogenesis. While there is evidence for a dual role of GLI1 in PDAC, this phenomenon has yet to be linked with other cancer types.”"
}
]
}
] | 1
|
https://f1000research.com/articles/3-238
|
https://f1000research.com/articles/5-1469/v1
|
22 Jun 16
|
{
"type": "Review",
"title": "Recent insights into the molecular mechanisms of the NLRP3 inflammasome activation",
"authors": [
"Tomasz Próchnicki",
"Matthew S. Mangan",
"Eicke Latz",
"Tomasz Próchnicki",
"Matthew S. Mangan"
],
"abstract": "Inflammasomes are high-molecular-weight protein complexes that are formed in the cytosolic compartment in response to danger- or pathogen-associated molecular patterns. These complexes enable activation of an inflammatory protease caspase-1, leading to a cell death process called pyroptosis and to proteolytic cleavage and release of pro-inflammatory cytokines interleukin (IL)-1β and IL-18. Along with caspase-1, inflammasome components include an adaptor protein, ASC, and a sensor protein, which triggers the inflammasome assembly in response to a danger signal. The inflammasome sensor proteins are pattern recognition receptors belonging either to the NOD-like receptor (NLR) or to the AIM2-like receptor family. While the molecular agonists that induce inflammasome formation by AIM2 and by several other NLRs have been identified, it is not well understood how the NLR family member NLRP3 is activated. Given that NLRP3 activation is relevant to a range of human pathological conditions, significant attempts are being made to elucidate the molecular mechanism of this process. In this review, we summarize the current knowledge on the molecular events that lead to activation of the NLRP3 inflammasome in response to a range of K+ efflux-inducing danger signals. We also comment on the reported involvement of cytosolic Ca2+ fluxes on NLRP3 activation. We outline the recent advances in research on the physiological and pharmacological mechanisms of regulation of NLRP3 responses, and we point to several open questions regarding the current model of NLRP3 activation.",
"keywords": [
"Inflammasome",
"pyroptosis",
"NLRP3",
"autoinflammatory disease"
],
"content": "Direct activation of NLRP3 by K+ efflux\n\nThe stimulatory effect that cytosolic K+ depletion has on IL-1β proteolytic processing and secretion from LPS-primed macrophages and monocytes, in response to compounds such as ATP or nigericin, was observed long before the discovery of inflammasomes1–3. This effect is now known to be mediated by NLRP34, and K+ efflux remains the best-characterized minimal stimulus for NLRP3 inflammasome activation5. Conversely, incubation in media containing supraphysiological [K+] can block NLRP3 inflammasome assembly in response to most of the identified NLRP3 triggers5.\n\nThe major classes of NLRP3 activators include extracellular ATP at millimolar concentrations, K+ ionophores4 and crystalline/particulate substances, or other factors that cause lysosomal destabilization6,7. All of these stimuli are known to decrease the cytosolic level of K+ ions5, but the mechanisms of K+ efflux induction (summarized in Figure 1) differ for all classes of NLRP3 triggers. Before discussing these mechanisms in detail, it is important to reiterate some of the basic features of cellular K+ homeostasis. Firstly, the cytosolic [K+] ([K+]i; ~140 mM) is much higher than the extracellular [K+] ([K+]e; ~5 mM), and this distribution is approximately reversed for Na+ ions. This asymmetry is maintained by the Na+/K+-ATPase, an electrogenic ion pump transporting, during each cycle, two K+ cations into the cytosol and three Na+ cations into the extracellular milieu. Secondly, under basal conditions, the permeability of most mammalian plasma membranes is highest with respect to the K+ cations and much lower for Na+ and Ca2+. Together, these factors contribute to sustaining the transmembrane potential of mammalian cells, which is characterized by a slight excess of negative charge on the inside of the cell. Each cycle of the Na+/K+-ATPase produces an electric charge difference of one elementary unit, and slow leakage of K+ ions from the cell is not counterbalanced by a compensatory influx of another type of cation8. Plasma membranes in basal states can generally be regarded as electrical insulators9, so translocations of even small numbers of individual ions (much too small to cause any measurable changes in the intracellular concentration of the respective ion) produce significant changes in the value of transmembrane potential8. Similarly, transporting ions in the direction opposite to the electrical gradient (i.e. cations to the outside of the cell or anions into the cell) requires significant energy input. In this light, the dramatic decrease in [K+]i required for NLRP3 activation, estimated as a drop of at least ~20–30%5, can be expected to be accompanied by either a counter-flux of cations or a “co-flux” of anions, which should also be provided by NLRP3 stimuli.\n\nUnder basal conditions, high intracellular K+ concentration is maintained by the activity of Na+/K+-ATPase, which actively imports K+ ions into the cell and generates an electrical gradient that favors movement of cations into the cytoplasm. Together with leak K+ channels, Na+/K+-ATPase contributes to the transmembrane potential, characterized by a slight excess of negative charges inside the cell. Under conditions of NLRP3 stimulation, this equilibrium is disturbed. ATP increases the open probability of P2X7R, a cation channel that allows for net exchange of intracellular K+ ions for extracellular Na+ or Ca2+ ions. This produces a net K+ efflux that acts as an NLRP3 activator. Activation of P2X7R is also accompanied by opening of pannexin-1 channels. During hypotonic stimulation, the regulatory volume decrease (RVD) response causes opening of K+ and Cl- channels, driving an efflux of K+ and Cl- ions to balance the intracellular and extracellular osmolarity values. To induce NLRP3 activation, this mechanism of K+ ions depletion additionally requires an influx of Ca2+ through TRP channels and activation of the kinase TAK1. NLRP3-activating K+ ionophores produce a net K+ efflux through different mechanisms. The peptide gramicidin can insert itself into plasma membranes, forming pores that are permeable to monovalent cations. This enables an exchange of intracellular K+ for extracellular Na+. Valinomycin, a neutral ionophore, is a cell-permeant compound that can bind to K+ ions, replacing the hydration shell of this cation. Consequently, K+ ions shielded by valinomycin molecules can pass across the plasma membrane without a requirement for opening a K+-permeable pore. Nigericin is a carboxylic ionophore that can bind to H+ or to K+. Both the H+- and K+-bound forms of nigericin are plasma membrane permeant. In this way, nigericin mediates K+ transport from the compartment with higher K+ concentration to the compartment with lower K+ concentration, concomitantly leading to a transient acidification of cytosol. In further stages, the increased cytosolic [H+] can stimulate Na+/H+ exchangers to extrude H+ ions from the cytosol, which is accompanied by Na+ influx90. Lysosomal damage caused by particulate materials or by other factors requires K+ efflux to induce NLRP3 activation, but it is unknown which factors are involved in this K+ depletion pathway.\n\nExtracellular ATP at millimolar concentrations acts as an agonist of a ligand-gated cation channel called P2X7 receptor (P2X7R), which is permissive to K+, Na+, and Ca2+10, allowing for cytosolic K+ efflux balanced by the influx of extracellular Na+ and Ca2+. K+ ionophores activating the NLRP3 inflammasome provide diverse pathways for K+ transport. Gramicidin, a peptide ionophore, allows for K+ efflux balanced by Na+ influx by inserting into plasma membranes to form monovalent cation (Na+/K+)-permissive pores11 in a manner electrochemically similar to the P2X7R. A different mechanism is employed by nigericin, a carboxylic ionophore that can exist in a free membrane-impermeant anionic form or as a neutral membrane-permeant complex when bound to a K+ cation or a proton (H+). In one electroneutral K+ efflux cycle mediated by nigericin, the ionophore anion binds to H+ on the outside of the cell, passes across the plasma membrane as nigericin-H, and releases the proton on the intracellular side. There, nigericin anion binds to K+, which is subsequently transported across the plasma membrane as nigericin-K and released on the outside of the cell12. This mechanism facilitates K+ efflux by allowing an H+ influx, leading to acidification of cytosol. Valinomycin, another NLRP3-activating K+ ionophore, also forms equimolar complexes with K+ but, unlike the neutral nigericin-K complexes, these complexes have a single positive charge (valinomycin-K+)12. Therefore, valinomycin-mediated K+ efflux is electrogenic and can only occur until the chemical gradient that is pushing K+ ions to the outside of the cell is balanced by the electric force drawing cations into the cell. It is currently unknown if such modest leakage of K+ ions could be sufficient to activate NLRP3, or if the valinomycin-mediated K+ efflux is accompanied by movement of another ionic species that would allow for a more pronounced decrease of [K+]i.\n\nIt has been demonstrated that cell treatment with crystalline/particulate stimuli, representing pathophysiologically relevant NLRP3 activators, also leads to depletion of intracellular K+5. However, the mechanism by which crystal-induced K+ efflux occurs is currently not understood. It seems plausible that, during lysosomal rupture caused by crystals, mixing of lysosomal lumina (low [K+], close to the extracellular concentration13) with the cytosolic contents could passively decrease [K+]i. However, the observation that the net K+ content of cells decreases upon treatment with crystals5 suggests, in the absence of data on the values of [K+]i, that plasma membrane-resident K+ channels or transporters might be involved in the crystal-elicited K+ efflux. One model of NLRP3 activation by monosodium urate (MSU) crystals proposes that, upon phagocytosis, in the acidic lysosomal environment, Na+ ions can be released from MSU particles leading to an increase in the osmolarity of the cell. This increase in osmolarity can be balanced by an influx of water from the extracellular space, which, it is suggested, dilutes [K+]i and thereby triggers NLRP3 activation14. While the proposed mechanism explains how MSU crystal-induced lysosomal damage can activate the NLRP3 inflammasome, it cannot account for NLRP3 activation with stimuli such as silica or cholesterol crystals because these particles do not dissociate in the lysosomal pH and are consequently not expected to influence cellular osmolarity. Investigating the kinetics and molecular mechanism of K+ loss in cells undergoing lysosomal damage may prove highly relevant for understanding how the different NLRP3 stimuli trigger inflammasome assembly and for designing treatments that specifically target NLRP3 activation by crystalline agents, which underlies multiple inflammatory diseases.\n\nA number of other conditions that deplete cytosolic K+ have been demonstrated to activate NLRP3. These include pharmacological inhibition of Na+/K+-ATPase3,5. Blocking Na+/K+-ATPase deprives cells of the mechanism maintaining both the asymmetric distribution of Na+ and K+ ions and the membrane potential, leading to a loss of K+ ions. Under conditions of Na+/K+-ATPase inhibition, K+ ions are no longer actively imported by the cell. Furthermore, dissipation of membrane potential, which accompanies Na+/K+-ATPase inhibition, eliminates the electrical force drawing K+ ions into the cell. A similar scenario can be predicted in the case where cells are incubated in a K+-free medium, which was also demonstrated to activate the NLRP3 inflammasome5, because K+-free media act as Na+/K+-ATPase inhibitors15. Finally, a distinct mechanism of K+ efflux is involved in NLRP3 activation by low-osmolarity media16. Here, the regulatory volume decrease (RVD) response leads to a concerted efflux of K+ and Cl- ions in an attempt to equilibrate intracellular and extracellular osmolarities17. In this particular case, however, K+ efflux does not seem to be sufficient for NLRP3 inflammasome activation, and an additional influx of Ca2+ ions into the cytosol is required16 (further discussed below).\n\n\nMolecular events occurring downstream of K+ efflux\n\nWhile depletion of intracellular K+ is required for NLRP3 activation, little is known about how the change in [K+]i is sensed and how this information is further transduced to the inflammasome. Recently, an important development has helped to solve this question, as NEK7, a Ser/Thr kinase involved in mitotic cell division, has been identified as a factor specifically required for NLRP3 inflammasome activation downstream of K+ efflux18–20. In response to NLRP3 activators, NEK7 is recruited to NLRP3 upstream of inflammasome formation (in a manner independent of ASC and caspases-1/11). NEK7 can also be detected in NLRP3/ASC specks, and formation of high-molecular-weight NLRP3 complexes that occurs upstream of ASC specking requires that NEK7 interacts with NLRP3. The NLRP3-NEK7 interaction is dependent on K+ efflux and can be blocked by high [K+]e18. Interestingly, the catalytic activity of NEK7 is required neither for its binding to NLRP3, nor for the activation of the NLRP3 inflammasome18,20.\n\nThe requirement for NEK7 in NLRP3 activation restricts this process to cells in interphase. At the endogenous level of NEK7, NLRP3 activators are able to enhance the interaction between NLRP3 and NEK7 in LPS-primed interphase cells but not in cells that have entered mitotic division, and LPS-primed interphase cells show a significantly higher level of caspase-1 activation than do their mitotic counterparts. Of note, NEK7 overexpression partially restores the responses to NLRP3 stimuli in mitotic cells, suggesting that the endogenous amount of NEK7 is not sufficient to simultaneously participate in both cell division and NLRP3 activation20.\n\nIt remains unknown how NEK7 is recruited to NLRP3 in response to a decrease in [K+]i and whether elevating the interaction between NLRP3 and NEK7 above a certain threshold is sufficient to trigger the NLRP3 inflammasome assembly. In partial response to the first question, it was found that stimulation with ATP increases NEK7 phosphorylation20. This increase could be blocked with N-acetylcysteine, a scavenger of reactive oxygen species (ROS) that also potently inhibits IL-1β release upon stimulation with ATP5,20,21. However, it has not been clearly demonstrated that the enhanced phosphorylation of NEK7 is required for its interaction with NLRP3 and for NLRP3 inflammasome activation. Furthermore, shRNA-mediated silencing of NEK9, a kinase that interacts with NEK7 and causes its activation22, does not inhibit NLRP3 activation18, suggesting either that NEK7 activation is not required for the NLRP3 inflammasome assembly or that the ROS-dependent NEK7 activation occurs through an as-yet-unidentified mechanism. Further elucidation of this discrepancy and of the detailed mechanism by which NEK7 contributes to NLRP3 activation will be the next important step towards understanding the molecular mechanism of inflammasome assembly.\n\nSeveral single amino acid substitutions in NLRP3 are causative for systemic inflammation observed in a spectrum of autoinflammatory diseases known as cryopyrin-associated periodic syndromes (CAPS; cryopyrin being a synonym of NLRP3): neonatal onset multisystem inflammatory disease (NOMID), Muckle-Wells syndrome (MWS), and familial cold autoinflammatory syndrome (FCAS)23. Of these, mainly the MWS-associated mutant NLRP3R260W (whose mouse counterpart is Nlrp3R258W) has been studied with respect to the requirement for K+ efflux for inflammasome assembly. Interestingly, activation of the Nlrp3R258W mutant occurs in macrophages expressing Nlrp3R258W in response to extracellular LPS stimulation (independent of any classical triggering stimuli) and without the requirement for K+ efflux5. However, NEK7 deficiency dramatically reduces the ability of Nlrp3R258W-expressing macrophages to activate the inflammasome in response to extracellular LPS18. NLRP3G775A and NLRP3G775R mutants, which are mainly associated with NOMID, show a stronger association with NEK7 than does WT-NLRP3 when overexpressed in HEK293 cells and, conversely, the inflammasome activation-incompetent NLRP3D946G mutant associates with NEK7 less strongly20. Collectively, these observations suggest that some of the CAPS-causative mutations in NLRP3 could promote inflammasome activation by facilitating the interaction between NLRP3 and NEK7, but such a conclusion requires further elucidation of the mechanism by which NEK7 is involved in the activation of different NLRP3 variants.\n\nMurine caspase-11 and its human orthologues caspases-4 and -5 are cytosolic LPS sensors24 that, upon recognition of their ligand, trigger non-canonical inflammasome activation25. This process consists of pyroptotic cell death that is independent of the canonical NLRP3 inflammasome components26 and NLRP3-, ASC-, and caspase-1-dependent IL-1β/IL-18 processing and secretion25. Recent studies demonstrated that NLRP3 activation downstream of caspase-11 is mediated by K+ efflux27,28. This process is initiated by caspase-11-mediated cleavage of pannexin-126, a plasma membrane-resident channel permeable to molecules and ions with a molecular weight of up to ~1 kDa29. Two molecular events follow the proteolytic processing of pannexin-1: (a) K+ efflux (a direct NLRP3 stimulus that induces mIL-1β secretion) and (b) release of ATP, which in turn acts as an agonist of P2X7R to promote cell death26. Surprisingly, the levels of ATP released from cells upon caspase-11 activation and proposed to activate P2X7R are much lower (nanomolar concentrations) than the amounts of ATP typically required to activate this receptor when added as an exogenous stimulus30. The mechanism by which intracellular LPS recognition increases macrophage sensitivity to extracellular ATP is not yet identified.\n\nAnother effector mechanism of caspase-11 activation involves proteolytic cleavage of a cytosolic protein, gasdermin D31–33. The signaling pathways activated upon cleavage of gasdermin D are unknown, but it was demonstrated that, while overexpression of full-length gasdermin D, or of gasdermin D C-terminal fragment, does not cause any apparent changes in cell physiology, expression of gasdermin D N-terminal fragment alone is highly cytotoxic32. This observation evinces that the N-terminal fragment of gasdermin D is one of the downstream effectors of caspase-11. Important questions to be answered in further investigation of the role of gasdermin D in non-canonical inflammasome activation are whether overexpression of the gasdermin D N-terminal fragment is sufficient to activate the NLRP3 inflammasome and whether this process involves K+ efflux. As gasdermin D N-terminal fragment alone is sufficient to cause pyroptosis, which is associated with plasma membrane disruption, it could be envisaged that this also leads to K+ depletion.\n\nSimilar to pannexin-126, gasdermin D is required for both cell death and IL-1β release in response to intracellular LPS32,33. While IL-1β secretion observed under conditions of intracellular LPS stimulation is dependent on NLRP3, pyroptosis elicited by intracellular LPS only depends on pannexin-126 and gasdermin D32,33 and is unaffected in NLRP3-deficient cells. The recent discoveries on the caspase-11 effector mechanisms leading to non-canonical inflammasome activation and to pyroptosis are summarized in Figure 2. Future studies should address the questions of whether—and how—the caspase-11-mediated events (pannexin-1 and gasdermin D proteolytic processing) converge to orchestrate pyroptotic cell death. In particular, the mechanism by which gasdermin D N-terminal fragment induces cytotoxicity will have to be resolved, and the potential role of pannexin-1 cleavage in this process will have to be investigated more closely. Furthermore, there is a proposed mechanism by which the caspase-11/pannexin-1/NLRP3 axis triggers IL-1β/IL-18 secretion26, but it remains unknown how gasdermin D is involved in this process.\n\nUpon recognition of LPS in the cytosol, caspase-11 cleaves pannexin-1 and gasdermin D. Cleavage of pannexin-1 leads to opening of the channel and leakage of K+ and ATP from the cell into the extracellular space. This efflux of K+ ions activates the NLRP3 inflammasome, causing proteolytic processing and secretion of IL-1β. Simultaneously, ATP acts as an agonist for the P2X7R, leading to NLRP3 inflammasome-independent pyroptotic cell death. Proteolytic cleavage of gasdermin D produces a highly toxic N-terminal fragment of this protein, which mediates both activation of the NLRP3 inflammasome (with subsequent IL-1β processing and secretion) and NLRP3-independent pyroptotic cell death. The relationship between two caspase-11 effectors, pannexin-1 and gasdermin D, is currently not understood.\n\nAdding to our knowledge on canonical inflammasome activation, active caspase-1, alongside caspase-11, was also demonstrated to cleave gasdermin D32. In response to activators of various canonical inflammasomes, gasdermin D-deficient macrophages exhibit delayed kinetics of cell death32 and decreased levels of secreted IL-1β31–33, which suggests that caspase-1-catalysed proteolysis of gasdermin D is one of the effector mechanisms of pyroptosis and that it may contribute to non-classical cytokine secretion.\n\nRecently, it was reported that inhibition of glycolysis by targeting enzymes that catalyze two of the late reactions of this pathway, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or α-enolase, is sufficient to elicit inflammasome activation in an NLRP3-dependent manner34. The proposed sequence of events consists of a decrease in cellular [NADH]/[NAD+] ratio, and a consequent increase in the levels of mitochondrial ROS, which, it is suggested, are involved in NLRP3 activation35,36. Interestingly, supplementation with the glycolytic metabolite pyruvate (which is normally produced downstream of the inhibited steps of glycolytic cascade) or with succinate (one of the TCA cycle metabolites; both pyruvate and succinate can enhance TCA cycle activity) leads to a decrease in the level of both generated mitochondrial ROS and NLRP3 inflammasome activation34. Such an observation suggests that in this pathway mitochondrial ROS act as NLRP3 stimuli. This mechanism of NLRP3 activation is uncommon because it does not require K+ efflux: inhibition of GAPDH and α-enolase can trigger assembly of the NLRP3 inflammasome at supraphysiological [K+]e34. Of note, NLRP3 activation under conditions of disrupted glycolytic flux could have profound pathophysiological significance as, for example, macrophages infected with Salmonella typhimurium exhibit decreased levels of NADH34. Given that infection with S. typhimurium is an activator of the NLRP3 inflammasome37 and that S. typhimurium-mediated NLRP3 activation can be abrogated by pyruvate supplementation34, this metabolic signature may constitute an important signal in inflammasome activation.\n\nOf note, it has been demonstrated that efficient glycolysis is required for NLRP3 activation by canonical K+-depleting stimuli38. Furthermore, the metabolic changes resulting from inhibition of glycolysis were not observed in cells treated with nigericin, a canonical K+ efflux-dependent NLRP3 activator34, suggesting that the newly discovered pathway is a distinct mechanism of NLRP3 activation rather than simply an event occurring downstream of cellular K+ depletion. In further support of this conclusion, canonical activation of the NLRP3 inflammasome with stimuli such as nigericin or ATP cannot be inhibited by supplementation with pyruvate34. It remains unknown whether the NLRP3 inflammasome assembly in response to glycolysis inhibitors relies on a NEK7-dependent mechanism. However, this is unlikely in light of the observations that (a) the interaction between NEK7 and NLRP3 requires K+ efflux and (b) the K+/H+ ionophore nigericin does not inflict metabolic changes resembling those caused by inhibition of glycolysis.\n\nAnother mechanism of NLRP3 activation independent of K+ efflux is observed in monocytes but is restricted to several species (e.g. humans or pigs) and not observed in murine cells. For this mechanism of NLRP3 activation, termed “alternative inflammasome activation”, extracellular LPS is a stimulus sufficient to elicit mature IL-1β release but not to cause ASC speck formation or pyroptosis. LPS acts as an agonist of TLR4, leading to engagement of the adaptor protein TRIF and of the RIPK1-FADD-caspase-8 signaling cascade, culminating in caspase-1 activation in an NLRP3- and ASC-dependent manner39. The specific nature of the signaling events that drive alternative NLRP3 activation as well as the fact that this mechanism does not lead to the generation of ASC specks may collectively suggest an involvement of a distinct, K+ efflux-independent active NLRP3 conformation in this process.\n\n\nCa2+ influx is not sufficient, and may not be required, for NLRP3 activation\n\nBased on the ability of certain NLRP3 stimuli to increase cytosolic [Ca2+] ([Ca2+]i), and on the inhibitory effect that several small-molecule compounds targeting intracellular Ca2+ have on NLRP3 activation, it was proposed that [Ca2+]i ions could be involved in NLRP3 activation40–42. The major mechanisms of [Ca2+]i increase in the cytosol are (a) Ca2+ influx from the lumen of endoplasmic reticulum (ER) through a ligand-gated ion channel called inositol trisphosphate (IP3) receptor (IP3R), a downstream effector of the phospholipase C (PLC) family, and (b) entry of extracellular Ca2+ ions through plasma membrane-resident Ca2+ channels43. Important Ca2+-buffering organelles are (c) mitochondria, which can either absorb or release Ca2+ under different conditions44. All of these pathways have been implicated in the activation of NLRP3. In favor of the hypothesis that the ER-derived Ca2+ ions could be a stimulus of NLRP3, a range of small-molecule IP3R antagonists and PLC inhibitors have been consistently demonstrated to inhibit activation of the NLRP3 inflammasome40–42. However, the observed levels of inhibition vary between the different studies, and in some cases the applied concentrations of small-molecule compounds required to inhibit NLRP3 significantly surpass their IC50 values reported for other processes45,46. Furthermore, artificially increasing [Ca2+]i with thapsigargin, an inhibitor of the sarcoplasmic/endoplasmic reticulum Ca2+-ATPase (SERCA; an ion pump transporting Ca2+ from the cytosol into the ER lumen, responsible for maintaining the steep [Ca2+] gradient between the ER lumen and the cytosol) either inhibits40 or does not influence47 NLRP3 activation, demonstrating that translocation of Ca2+ ions into the cytosol is not sufficient to trigger that process. Of note, thapsigargin was demonstrated to elicit modest IL-1β secretion from LPS-primed human macrophages, but it is not known whether the mature form of the cytokine is secreted and whether this process is mediated by NLRP348. Furthermore, thapsigargin was demonstrated to cause NLRP3 activation by inducing ER stress, but the role of Ca2+ in this process has not been studied49.\n\nThe current evidence for the involvement of extracellular Ca2+ in the activation of NLRP3 strongly suggests that this pool of Ca2+ does not play a role in the inflammasome assembly. Supporting this is the observation that all tested canonical NLRP3 stimuli that act by depleting cytosolic K+ can activate the inflammasome in Ca2+-free extracellular buffers47,50. In several studies, a contradictory effect of extracellular Ca2+ depletion was reported40,48 but, in some cases at least, such observations may have resulted from simultaneous application of BAPTA-AM and Ca2+-free buffer48, which can be expected to interfere with Ca2+ fluxes deriving from a range of different sources. Nevertheless, there is currently no convincing explanation as to why in certain experimental systems the removal of extracellular Ca2+ seems to inhibit NLRP3, while in other setups such treatment does not interfere with NLRP3 activation. A second argument supporting the claim that extracellular Ca2+ is not required for the NLRP3 inflammasome activation comes from the observation that K+ ionophores elicit NLRP3 inflammasome assembly in the absence of any significant changes in [Ca2+]i47. Surprisingly, one study reports that human macrophages can secrete IL-1β in response to ionomycin, a Ca2+ ionophore, but, similar to the case of thapsigargin, the involvement of NLRP3 is not proven and it is not demonstrated that the released form of the cytokine is proteolytically processed48.\n\nInterestingly, in the NLRP3 response to hypotonic environments, extracellular Ca2+ influx through mechanosensitive TRP channels and consequent activation of the kinase TAK1 were both demonstrated to be required for activation of the inflammasome alongside RVD-mediated K+ efflux16. It is unknown whether this mechanism also contributes to inflammasome assembly by all NLRP3 stimuli, but it is suggested that TAK1 is involved in NLRP3 activation induced by lysosomal damage51. Finally, one study proposed that Ca2+ influx into the cytosol follows K+ efflux caused by NLRP3 stimuli and that this Ca2+ flux promotes activation of the inflammasome by enhancing mitochondrial ROS generation52. However, there are two major limitations to this conclusion. First, the only method applied for targeting cytosolic Ca2+ was cell loading with BAPTA-AM, which may exert numerous off-target effects53. Secondly, the only tested NLRP3 activator was ATP, a cation channel opener and inducer of PLC (through interaction with P2Y2 receptor54), which makes it challenging to dissect the relative contributions of the various ion fluxes in the process of inflammasome activation.\n\nThe mitochondrial Ca2+ stores are the most difficult to perturb experimentally, and consequently their potential role in NLRP3 activation is not well understood. Several reports suggest that an increased Ca2+ uptake by the mitochondria may promote mitochondrial damage and NLRP3 responses40,55,56, although it was also demonstrated that mitochondrial damage inflicted by canonical NLRP3 activators is at least in part dependent on NLRP3 and caspase-157. The transporter responsible for mitochondrial Ca2+ uptake during NLRP3 activation by the membrane attack complex (a component of the complement cascade) and by Pseudomonas aeruginosa has been identified as mitochondrial Ca2+ uniporter (MCU)55,56. The specific factors that could tie an increase in mitochondrial [Ca2+] to NLRP3 inflammasome assembly have not been identified, and it remains unknown whether MCU is involved in NLRP3 responses to its classical, better-characterized activators.\n\nThe collective evidence regarding the role of Ca2+ in the activation of NLRP3 suggests that elevation in [Ca2+]i is not required for the assembly of this inflammasome. However, a modulatory role for this ion cannot be excluded, especially in light of two puzzling observations: (a) inhibition of NLRP3 responses by BAPTA-AM41,47 and (b) inhibition of NLRP3 activation by siRNA knock-down of a Gqα-coupled G-protein-coupled receptor (GPCR) called Ca2+-sensing receptor (CaSR)42. Given that BAPTA-AM has off-target effects apart from scavenging Ca2+ ions inside the cell53 and that Ca2+ ions are not the only ligand of CaSR58, re-evaluation of the mechanisms by which application of BAPTA-AM or suppression of CaSR signaling interfere with NLRP3 activation could provide valuable insights into the molecular events that regulate NLRP3 inflammasome assembly. We further discuss some aspects of GPCR/CaSR signaling below.\n\n\nPhysiological and pharmacological modulation of NLRP3\n\nThe relevance of NLRP3 in human pathologies has led to research regarding both the intrinsic mechanisms that limit inflammasome activation and the possibility of pharmacological targeting of NLRP3. Even though such studies are challenging, because it is unclear how the K+ efflux is transduced to NLRP3, they have nevertheless resulted in discoveries that processes such as cAMP signaling and autophagy can interfere with NLRP3 activation and in identification of several classes of exogenous small-molecule compounds that can act as specific inhibitors of NLRP3 activation.\n\nIn recent years, the interest in how GPCRs could regulate NLRP3 responses resulted in an observation that increasing [cAMP]i inhibits the activation of the NLRP3 inflammasome42. Specifically, treating cells with pharmacological activators of adenylyl cyclases42, or with agonists of GPCRs that enhance adenylyl cyclase activity59,60, leads to a decrease in NLRP3 activation in response to classical NLRP3 stimuli. Conversely, NLRP3 stimuli were demonstrated to decrease [cAMP]i42, although it is currently not clear whether this decrease occurs upstream or downstream of NLRP3 activation. One study suggested that inhibition of adenylyl cyclase enzymatic activity (surprisingly, using KH7, an inhibitor targeting the GPCR-independent soluble adenylyl cyclase and not acting on the GPCR-regulated transmembrane adenylyl cyclases61) might be sufficient to activate the NLRP3 inflammasome42, but this result could not be reproduced, possibly due to differences in the applied concentrations of the compound60. Of note, inhibitors of transmembrane adenylyl cyclases also do not act as NLRP3 inflammasome activators, pointing to a modulatory role of [cAMP]i rather than its decrease being the direct NLRP3 stimulus.\n\nPharmacological targeting of various cAMP-binding proteins that act as downstream effectors of adenylyl cyclase activation revealed that the inhibitory effect that cAMP exerts on NLRP3 activation cannot be ascribed to the currently known cAMP targets42,59,60. In the study that identified cAMP as a regulator of NLRP3, it was proposed that NLRP3 could form a complex with cAMP42, and recently it was demonstrated that the NLRP3-cAMP complex recruits ubiquitin ligase MARCH7, which in turn labels NLRP3 for degradation in autophagosomes60. It is suggested that this process down-regulates NLRP3 signaling. Nevertheless, there are still open questions about the mechanism of inhibition of NLRP3 responses by cAMP. This model suggests a direct interaction between cAMP and NLRP3, in which the nucleotide-binding domain (NBD) of NLRP3 is involved42. However, sequence analysis of NLRP3-NBD does not suggest the presence of a cyclic nucleotide-binding fold along with the ATP-binding site62. Furthermore, even if the cAMP-binding and ATP-binding sites were in fact the same structural interface, it is not likely that cAMP could compete with ATP for binding to NLRP3, given that in living cells [ATP]i63,64 is much higher than [cAMP]i65. Another problematic aspect of studies on the influence of cAMP on activation of NLRP3 is the consistent use of KH759,60, the inhibitor of soluble, GPCR-independent adenylyl cyclases66, to interfere with events that, it is proposed, occur downstream of GPCR-responsive transmembrane adenylyl cyclases. Applying genetic rather than pharmacological approaches to the studies on the influence of cAMP on NLRP3 activation and a more thorough investigation of the roles of established cAMP-binding proteins in this process could potentially provide a greater insight into the mechanism of NLRP3 response inhibition by cAMP.\n\nAutophagy is emerging as a central process regulating multiple inflammasome responses at several levels. Pro-IL-1β can be degraded in autophagosomes, leading to decreased inflammatory responses to a range of stimuli67. In addition, it is also proposed that autophagy specifically controls NLRP3 activation over other characterized inflammasomes. Suppression of the autophagic processes impairs homeostatic turnover of mitochondria, promoting mitochondrial damage that contributes to caspase-1 activation in response to ATP68, as well as in the NLRP3 response to influenza A virus infection69. A decrease in the number of autophagosomes was also reported in response to palmitate, a long-chain fatty acid previously demonstrated to activate NLRP370. However, activation of NLRP3 using nigericin or crystalline stimuli enhanced autophagy,71 targeting inflammasome components for degradation in the autophagosomes. Collectively, these observations suggest that reduction in the autophagic processing of cellular contents may support NLRP3 inflammasome responses. Conversely, increased autophagy may act as a regulator of the NLRP3 inflammasome specifically and a regulator of IL-1β-based inflammation generally by a negative feedback loop. The recent discoveries that dopamine decreases cellular responses to NLRP3 activators by targeting this inflammasome sensor protein for degradation in autophagosomes60 and that NF-κB signaling can inhibit activation of NLRP3 by stimulating the autophagic turnover of dysfunctional mitochondria72 demonstrate that this regulatory mechanism can position immune cells towards a state of decreased sensitivity to NLRP3 stimuli even before they encounter inflammasome activators. The proposed mechanisms of regulation of NLRP3 responses by autophagy and by cAMP (discussed earlier) are summarized in Figure 3.\n\nSeveral physiological mechanisms regulate NLRP3 responses on the cellular level. Agonists of GS-coupled GPCRs stimulate the generation of cAMP by transmembrane adenylyl cyclases. cAMP is believed to bind to the nucleotide-binding domain of NLRP3. This formed NLRP3-cAMP complex recruits the ubiquitin ligase MARCH7 that polyubiquitinates NLRP3, targeting it for autophagosomal degradation. Autophagosomes are also the organelles responsible for degradation of pro-IL-1β (the inactive pro-form of the proinflammatory cytokine IL-1β), which is a more general mechanism controlling the inflammatory responses mediated by a range of inflammasomes. Finally, mitophagy is a way to dispose of damaged mitochondria that starts with sequestering them in autophagosomes. Autophagosomal degradation of dysfunctional mitochondria curbs the inflammasome responses, possibly by removing the source of direct NLRP3 activators.\n\nThe first observation that compounds containing a sulfonylurea moiety potently inhibit ATP- or hypotonicity-induced IL-1β processing and release predates the discovery of inflammasomes73. This phenomenon was later recognized as specific inhibition of the NLRP3 inflammasome74, and until now virtually all validated NLRP3 activators are sensitive to sulfonylurea-containing compounds, such as glyburide or CP-456,77375. Sulfonylurea drugs seem to specifically inhibit the triggering step of NLRP3 activation without affecting the NF-κB signaling-related priming step or the activation of other inflammasomes74,75. Compounds containing sulfonylurea moieties have been tested, as NLRP3 inhibitors, in several animal inflammatory disease models, usually with encouraging results75–79. Of note, alternative inflammasome activation (described in more detail above) can also be blocked by CP-456,77339.\n\nThe mechanism by which sulfonylurea compounds inhibit NLRP3 activation is currently not understood. Given that an important target of these pharmaceuticals are K+ channels80–82 and that K+ efflux is required for NLRP3 activation5, one concept would be that sulfonylureas could impede K+ efflux from cells treated with NLRP3 stimuli. However, not all sulfonylurea drugs can inhibit inflammasome activation74 and, conversely, sulfonylurea compounds were demonstrated not to prevent K+ efflux caused by NLRP3 activators75, which collectively suggests that these inhibitors act downstream of K+ depletion and that the inhibition mechanism is not related to the activity of these compounds on K+ channels. Glyburide was shown to inhibit the ATPase activity of NLRP3, but it is unclear whether other drugs from this class can act in a similar manner, and if this observation is related to the glyburide-mediated inhibition of inflammasome formation. CP-456,773 (CRID383, which has recently been renamed to MCC95075) has been demonstrated not to affect the Ca2+ flux in cells treated with ATP75, which, some studies suggest, plays a role in NLRP3 activation40. The influence of CP-456,773 on other molecular events connected to NLRP3 activation, such as the production of ROS, decrease in [cAMP]i, or recruitment of NEK7 to NLRP3, has not yet been tested.\n\nAttempts to identify the molecular target of CP-456,773 showed that this compound interacts with proteins from the glutathione S-transferase family83, but so far none of these have been shown to transduce the information about K+ efflux to the NLRP3 inflammasome. There is conflicting evidence regarding the ability of sulfonylurea drugs to inhibit the activation of CAPS-related NLRP3 mutants, as glyburide has been shown not to affect IL-1β release from cultured monocytes from an FCAS-affected patient74, but CP-456,773 suppressed mutant NLRP3 activation in both the mouse model of MWS and in monocytes from an MWS-affected patient75. The NLRP3 mutants investigated in these studies had different amino acid substitutions, which, together with other differences in the experimental systems, could have led to this apparent discrepancy. A more comprehensive study addressing the sensitivity of a range of hyperactive NLRP3 mutants to sulfonylurea compounds could provide more insight into whether these drugs can inhibit inflammasome activation by these protein variants and what the mechanism of inhibition could be.\n\nSeveral recent studies demonstrated that NLRP3 activation can be abolished by pre-treatment of cells with various compounds that contain a Michael acceptor group (a double C=C bond in the α position with respect to a carbonyl [-C=O] or a nitro group [-NO2])84,85. In biological systems, these compounds can covalently modify protein Cys residues that are not engaged in disulfide bond formation86,87. This mechanism explains the inhibition of the NLRP3 inflammasome both by compounds that are generally regarded as NLRP3 inhibitors (e.g. parthenolide and BAY 11-7082)85 and compounds whose ability to inhibit NLRP3 activation has been identified as an “off-target” effect (e.g. the Syk kinase inhibitor 3,4-methylenedioxy-β-nitrostyrene [MNS])84. Importantly, the issue of specificity of the tested inflammasome inhibitors has also been addressed in the cited studies and, while MNS and BAY 11-7082 have been demonstrated to selectively inhibit NLRP3, parthenolide was also able to block other inflammasome responses84,85. This observation is probably related to parthenolide-mediated direct inhibition of caspase-185 (which contains a Cys residue that is essential for its catalytic activity and which can also be modified by Michael acceptors87).\n\nIn further investigation of the mechanism of NLRP3 inhibition by Cys-modifying compounds, two consecutive structure-activity relationship studies demonstrated that Michael acceptors with very diverse chemical structures can interfere with NLRP3 activation (assessed by IL-1β release and pyroptotic LDH release)88,89. Of note, these compounds are capable of inhibiting NLRP3 ATPase activity88,89. Furthermore, several Michael acceptors moderately but significantly inhibit the activation of CAPS-related NLRP3 variants, but their potency on these NLRP3 mutants is lower compared to the inhibitory influence exerted on WT-NLRP389.\n\nWhen applying compounds that contain a Michael acceptor moiety to investigate the molecular mechanism of NLRP3 activation, several issues have to be considered. First, the downstream effector of NLRP3 is caspase-1, a Cys protease whose catalytic activity depends on an unmodified, free Cys residue. This implies that Michael acceptors, at high enough concentrations, may obscure various experimental readouts that rely on caspase-1 activity, such as assessing inflammasome speck formation using the caspase-1-targeting FLICA reagent, IL-1β/IL-18 release, or pyroptotic LDH release, even if the particular compounds-of-interest do not directly interfere with NLRP3 activation. Second, the ability of Michael acceptors to directly interact with NLRP3/modify Cys residues in NLRP384,89 only suggests, but does not prove, that the NLRP3-Cys modification constitutes the mechanism of inflammasome inhibition by these compounds. Further insights into the structural basis of NLRP3 activation and into the possible influence of Michael acceptors on that process are required to resolve this issue.\n\n\nConclusions and future directions\n\nIn recent years, a number of molecular players involved in NLRP3 activation have been identified. Most importantly, the direct interaction of NLRP3 with the kinase NEK7 has been described, and its importance for the assembly of the NLRP3 inflammasome has been demonstrated. Major progress has also been made in our understanding of how intracellular LPS triggers caspase-11, leading to proteolytic processing of pannexin-1 and gasdermin D, and to non-canonical NLRP3 inflammasome activation. Finally, the mechanism by which human monocytes activate NLRP3 in response to extracellular LPS as a single stimulus has been solved and shown to rely on TLR4/TRIF-mediated activation of the RIPK1-FADD-caspase-8 cascade. On the other hand, physiologically relevant mechanisms, such as autophagy and cAMP signaling, have been proposed to down-regulate the activation of NLRP3, demonstrating the physiological importance of limiting NLRP3 inflammasome responses.\n\nHowever, there still remain unanswered questions about the molecular events that link cytosolic K+ depletion to NLRP3-/NEK7-dependent inflammasome formation. Furthermore, the discovery of the dual function of gasdermin D (as a downstream effector of caspase-11 required for non-canonical NLRP3 activation and a substrate of inflammatory caspases required for pyroptosis) calls for closer investigation of the mechanism of action of this protein. In light of the reported K+ efflux-independent modes of NLRP3 triggering that include alternative NLRP3 inflammasome activation and NLRP3 inflammasome activation upon inhibition of glycolysis, the relative contributions of these pathways to inflammatory responses will have to be evaluated. Finally, the mechanisms by which compounds containing sulfonylurea or Michael acceptor moieties cause NLRP3 inhibition will have to be defined, which may open new possibilities for potential future therapeutic applications of these molecules.",
"appendix": "Author contributions\n\n\n\nTP wrote the manuscript and prepared the figures. MSM edited the manuscript and provided valuable discussions and criticism. TP, MSM, and EL critically read, analyzed, and discussed the primary sources and conceived the outline of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe would like to thank Dr Dagmar Wachten, the Leader of the Max Planck Research Group Molecular Physiology, for helpful discussions on the signaling mechanisms involved in activation of the NLRP3 inflammasome.\n\n\nReferences\n\nPerregaux D, Barberia J, Lanzetti AJ, et al.: IL-1 beta maturation: evidence that mature cytokine formation can be induced specifically by nigericin. J Immunol. 1992; 149(4): 1294–303. PubMed Abstract\n\nPerregaux D, Gabel CA: Interleukin-1 beta maturation and release in response to ATP and nigericin. Evidence that potassium depletion mediated by these agents is a necessary and common feature of their activity. J Biol Chem. 1994; 269(21): 15195–203. PubMed Abstract\n\nWalev I, Reske K, Palmer M, et al.: Potassium-inhibited processing of IL-1 beta in human monocytes. EMBO J. 1995; 14(8): 1607–14. PubMed Abstract | Free Full Text\n\nMariathasan S, Weiss DS, Newton K, et al.: Cryopyrin activates the inflammasome in response to toxins and ATP. Nature. 2006; 440(7081): 228–32. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMuñoz-Planillo R, Kuffa P, Martínez-Colón G, et al.: K+ efflux is the common trigger of NLRP3 inflammasome activation by bacterial toxins and particulate matter. Immunity. 2013; 38(6): 1142–53. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nMartinon F, Pétrilli V, Mayor A, et al.: Gout-associated uric acid crystals activate the NALP3 inflammasome. Nature. 2006; 440(7081): 237–41. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nHornung V, Bauernfeind F, Halle A, et al.: Silica crystals and aluminum salts activate the NALP3 inflammasome through phagosomal destabilization. Nat Immunol. 2008; 9(8): 847–56. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWright SH: Generation of resting membrane potential. Adv Physiol Educ. 2004; 28(1–4): 139–42. PubMed Abstract | Publisher Full Text\n\nNiebur E: Electrical properties of cell membranes. Scholarpedia. 2008; 3(6): 7166. Publisher Full Text\n\nAlves LA, de Melo Reis RA, de Souza CA, et al.: The P2X7 receptor: shifting from a low- to a high-conductance channel - an enigmatic phenomenon? Biochim Biophys Acta. 2014; 1838(10): 2578–87. PubMed Abstract | Publisher Full Text\n\nKelkar DA, Chattopadhyay A: The gramicidin ion channel: a model membrane protein. Biochim Biophys Acta. 2007; 1768(9): 2011–25. PubMed Abstract | Publisher Full Text\n\nPressman BC: Biological applications of ionophores. Annu Rev Biochem. 1976; 45: 501–30. PubMed Abstract | Publisher Full Text\n\nDeng D, Jiang N, Hao SJ, et al.: Loss of membrane cholesterol influences lysosomal permeability to potassium ions and protons. Biochim Biophys Acta. 2009; 1788(2): 470–6. PubMed Abstract | Publisher Full Text\n\nSchorn C, Frey B, Lauber K, et al.: Sodium overload and water influx activate the NALP3 inflammasome. J Biol Chem. 2011; 286(1): 35–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChurchill PC: Second messengers in renin secretion. Am J Physiol. 1985; 249(2 Pt 2): F175–84. PubMed Abstract\n\nCompan V, Baroja-Mazo A, López-Castejón G, et al.: Cell volume regulation modulates NLRP3 inflammasome activation. Immunity. 2012; 37(3): 487–500. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nOrdaz B, Tuz K, Ochoa LD, et al.: Osmolytes and mechanisms involved in regulatory volume decrease under conditions of sudden or gradual osmolarity decrease. Neurochem Res. 2004; 29(1): 65–72. PubMed Abstract | Publisher Full Text\n\nHe Y, Zeng MY, Yang D, et al.: NEK7 is an essential mediator of NLRP3 activation downstream of potassium efflux. Nature. 2016; 530(7590): 354–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSchmid-Burgk JL, Chauhan D, Schmidt T, et al.: A Genome-wide CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) Screen Identifies NEK7 as an Essential Component of NLRP3 Inflammasome Activation. J Biol Chem. 2016; 291(1): 103–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nShi H, Wang Y, Li X, et al.: NLRP3 activation and mitosis are mutually exclusive events coordinated by NEK7, a new inflammasome component. Nat Immunol. 2016; 17(3): 250–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nJuliana C, Fernandes-Alnemri T, Kang S, et al.: Non-transcriptional priming and deubiquitination regulate NLRP3 inflammasome activation. J Biol Chem. 2012; 287(43): 36617–22. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHaq T, Richards MW, Burgess SG, et al.: Mechanistic basis of Nek7 activation through Nek9 binding and induced dimerization. Nat Commun. 2015; 6: 8771. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBroderick L, De Nardo D, Franklin BS, et al.: The inflammasomes and autoinflammatory syndromes. Annu Rev Pathol. 2015; 10: 395–424. PubMed Abstract | Publisher Full Text\n\nShi J, Zhao Y, Wang Y, et al.: Inflammatory caspases are innate immune receptors for intracellular LPS. Nature. 2014; 514(7521): 187–92. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKayagaki N, Warming S, Lamkanfi M, et al.: Non-canonical inflammasome activation targets caspase-11. Nature. 2011; 479(7371): 117–21. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nYang D, He Y, Muñoz-Planillo R, et al.: Caspase-11 Requires the Pannexin-1 Channel and the Purinergic P2X7 Pore to Mediate Pyroptosis and Endotoxic Shock. Immunity. 2015; 43(5): 923–32. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nSchmid-Burgk JL, Gaidt MM, Schmidt T, et al.: Caspase-4 mediates non-canonical activation of the NLRP3 inflammasome in human myeloid cells. Eur J Immunol. 2015; 45(10): 2911–7. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nRühl S, Broz P: Caspase-11 activates a canonical NLRP3 inflammasome by promoting K+ efflux. Eur J Immunol. 2015; 45(10): 2927–36. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nPelegrin P, Surprenant A: Pannexin-1 mediates large pore formation and interleukin-1beta release by the ATP-gated P2X7 receptor. EMBO J. 2006; 25(21): 5071–82. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBartlett R, Stokes L, Sluyter R: The P2X7 receptor channel: recent developments and the use of P2X7 antagonists in models of disease. Pharmacol Rev. 2014; 66(3): 638–75. PubMed Abstract | Publisher Full Text\n\nHe W, Wan H, Hu L, et al.: Gasdermin D is an executor of pyroptosis and required for interleukin-1β secretion. Cell Res. 2015; 25(12): 1285–98. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nKayagaki N, Stowe IB, Lee BL, et al.: Caspase-11 cleaves gasdermin D for non-canonical inflammasome signalling. Nature. 2015; 526(7575): 666–71. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nShi J, Zhao Y, Wang K, et al.: Cleavage of GSDMD by inflammatory caspases determines pyroptotic cell death. Nature. 2015; 526(7575): 660–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSanman LE, Qian Y, Eisele NA, et al.: Disruption of glycolytic flux is a signal for inflammasome signaling and pyroptotic cell death. eLife. 2016; 5: pii: e13663. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nHeid ME, Keyel PA, Kamga C, et al.: Mitochondrial reactive oxygen species induces NLRP3-dependent lysosomal damage and inflammasome activation. J Immunol. 2013; 191(10): 5230–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou R, Yazdi AS, Menu P, et al.: A role for mitochondria in NLRP3 inflammasome activation. Nature. 2011; 469(7329): 221–5. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBroz P, Newton K, Lamkanfi M, et al.: Redundant roles for inflammasome receptors NLRP3 and NLRC4 in host defense against Salmonella. J Exp Med. 2010; 207(8): 1745–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoon JS, Hisata S, Park MA, et al.: mTORC1-Induced HK1-Dependent Glycolysis Regulates NLRP3 Inflammasome Activation. Cell Rep. 2015; 12(1): 102–15. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nGaidt MM, Ebert TS, Chauhan D, et al.: Human Monocytes Engage an Alternative Inflammasome Pathway. Immunity. 2016; 44(4): 833–46. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nMurakami T, Ockinger J, Yu J, et al.: Critical role for calcium mobilization in activation of the NLRP3 inflammasome. Proc Natl Acad Sci U S A. 2012; 109(28): 11282–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRossol M, Pierer M, Raulien N, et al.: Extracellular Ca2+ is a danger signal activating the NLRP3 inflammasome through G protein-coupled calcium sensing receptors. Nat Commun. 2012; 3: 1329. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee G, Subramanian N, Kim AI, et al.: The calcium-sensing receptor regulates the NLRP3 inflammasome through Ca2+ and cAMP. Nature. 2012; 492(7427): 123–7. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nClapham DE: Calcium signaling. Cell. 2007; 131(6): 1047–58. PubMed Abstract | Publisher Full Text\n\nWalsh C, Barrow S, Voronina S, et al.: Modulation of calcium signalling by mitochondria. Biochim Biophys Acta. 2009; 1787(11): 1374–82. PubMed Abstract | Publisher Full Text\n\nBleasdale JE, Thakur NR, Gremban RS, et al.: Selective inhibition of receptor-coupled phospholipase C-dependent processes in human platelets and polymorphonuclear neutrophils. J Pharmacol Exp Ther. 1990; 255(2): 756–68. PubMed Abstract\n\nMaruyama T, Kanaji T, Nakade S, et al.: 2APB, 2-aminoethoxydiphenyl borate, a membrane-penetrable modulator of Ins(1,4,5)P3-induced Ca2+ release. J Biochem. 1997; 122(3): 498–505. PubMed Abstract | Publisher Full Text\n\nKatsnelson MA, Rucker LG, Russo HM, et al.: K+ efflux agonists induce NLRP3 inflammasome activation independently of Ca2+ signaling. J Immunol. 2015; 194(8): 3937–52. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nRada B, Park JJ, Sil P, et al.: NLRP3 inflammasome activation and interleukin-1β release in macrophages require calcium but are independent of calcium-activated NADPH oxidases. Inflamm Res. 2014; 63(10): 821–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMenu P, Mayor A, Zhou R, et al.: ER stress activates the NLRP3 inflammasome via an UPR-independent pathway. Cell Death Dis. 2012; 3: e261. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrough D, Le Feuvre RA, Wheeler RD, et al.: Ca2+ stores and Ca2+ entry differentially contribute to the release of IL-1 beta and IL-1 alpha from murine macrophages. J Immunol. 2003; 170(6): 3029–36. PubMed Abstract | Publisher Full Text\n\nOkada M, Matsuzawa A, Yoshimura A, et al.: The lysosome rupture-activated TAK1-JNK pathway regulates NLRP3 inflammasome activation. J Biol Chem. 2014; 289(47): 32926–36. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYaron JR, Gangaraju S, Rao MY, et al.: K+ regulates Ca2+ to drive inflammasome signaling: dynamic visualization of ion flux in live cells. Cell Death Dis. 2015; 6: e1954. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSaoudi Y, Rousseau B, Doussière J, et al.: Calcium-independent cytoskeleton disassembly induced by BAPTA. Eur J Biochem. 2004; 271(15): 3255–64. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nErb L, Weisman GA: Coupling of P2Y receptors to G proteins and other signaling pathways. Wiley Interdiscip Rev Membr Transp Signal. 2012; 1(6): 789–803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTriantafilou K, Hughes TR, Triantafilou M, et al.: The complement membrane attack complex triggers intracellular Ca2+ fluxes leading to NLRP3 inflammasome activation. J Cell Sci. 2013; 126(Pt 13): 2903–13. PubMed Abstract | Publisher Full Text\n\nRimessi A, Bezzerri V, Patergnani S, et al.: Mitochondrial Ca2+-dependent NLRP3 activation exacerbates the Pseudomonas aeruginosa-driven inflammatory response in cystic fibrosis. Nat Commun. 2015; 6: 6201. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nYu J, Nagasu H, Murakami T, et al.: Inflammasome activation leads to Caspase-1-dependent mitochondrial damage and block of mitophagy. Proc Natl Acad Sci U S A. 2014; 111(43): 15514–9. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nBrennan SC, Thiem U, Roth S, et al.: Calcium sensing receptor signalling in physiology and cancer. Biochim Biophys Acta. 2013; 1833(7): 1732–44. PubMed Abstract | Publisher Full Text\n\nSokolowska M, Chen L, Liu Y, et al.: Prostaglandin E2 Inhibits NLRP3 Inflammasome Activation through EP4 Receptor and Intracellular Cyclic AMP in Human Macrophages. J Immunol. 2015; 194(11): 5472–87. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYan Y, Jiang W, Liu L, et al.: Dopamine controls systemic inflammation through inhibition of NLRP3 inflammasome. Cell. 2015; 160(1–2): 62–73. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nBitterman JL, Ramos-Espiritu L, Diaz A, et al.: Pharmacological distinction between soluble and transmembrane adenylyl cyclases. J Pharmacol Exp Ther. 2013; 347(3): 589–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRehmann H, Wittinghofer A, Bos JL: Capturing cyclic nucleotides in action: snapshots from crystallographic studies. Nat Rev Mol Cell Biol. 2007; 8(1): 63–73. PubMed Abstract | Publisher Full Text\n\nGribble FM, Loussouarn G, Tucker SJ, et al.: A novel method for measurement of submembrane ATP concentration. J Biol Chem. 2000; 275(39): 30046–9. PubMed Abstract | Publisher Full Text\n\nLarcombe-McDouall J, Buttell N, Harrison N, et al.: In vivo pH and metabolite changes during a single contraction in rat uterine smooth muscle. J Physiol. 1999; 518(Pt 3): 783–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIancu RV, Ramamurthy G, Warrier S, et al.: Cytoplasmic cAMP concentrations in intact cardiac myocytes. Am J Physiol Cell Physiol. 2008; 295(2): C414–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRahman N, Buck J, Levin LR: pH sensing via bicarbonate-regulated \"soluble\" adenylyl cyclase (sAC). Front Physiol. 2013; 4: 343. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris J, Hartman M, Roche C, et al.: Autophagy controls IL-1beta secretion by targeting pro-IL-1beta for degradation. J Biol Chem. 2011; 286(11): 9587–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakahira K, Haspel JA, Rathinam VAK, et al.: Autophagy proteins regulate innate immune responses by inhibiting the release of mitochondrial DNA mediated by the NALP3 inflammasome. Nat Immunol. 2011; 12(3): 222–30. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nLupfer C, Thomas PG, Anand PK, et al.: Receptor interacting protein kinase 2-mediated mitophagy regulates inflammasome activation during virus infection. Nat Immunol. 2013; 14(5): 480–8. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nWen H, Gris D, Lei Y, et al.: Fatty acid-induced NLRP3-ASC inflammasome activation interferes with insulin signaling. Nat Immunol. 2011; 12(5): 408–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShi C, Shenderov K, Huang N, et al.: Activation of autophagy by inflammatory signals limits IL-1β production by targeting ubiquitinated inflammasomes for destruction. Nat Immunol. 2012; 13(3): 255–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhong Z, Umemura A, Sanchez-Lopez E, et al.: NF-κB Restricts Inflammasome Activation via Elimination of Damaged Mitochondria. Cell. 2016; 164(5): 896–910. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nPerregaux DG, McNiff P, Laliberte R, et al.: Identification and characterization of a novel class of interleukin-1 post-translational processing inhibitors. J Pharmacol Exp Ther. 2001; 299(1): 187–97. PubMed Abstract\n\nLamkanfi M, Mueller JL, Vitari AC, et al.: Glyburide inhibits the Cryopyrin/Nalp3 inflammasome. J Cell Biol. 2009; 187(1): 61–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nColl RC, Robertson AAB, Chae JJ, et al.: A small-molecule inhibitor of the NLRP3 inflammasome for the treatment of inflammatory diseases. Nat Med. 2015; 21(3): 248–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao J, Kapadia VS, Brown LS, et al.: The NLRP3 inflammasome is critically involved in the development of bronchopulmonary dysplasia. Nat Commun. 2015; 6: 8977. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nKrishnan SM, Dowling JK, Ling YH, et al.: Inflammasome activity is essential for one kidney/deoxycorticosterone acetate/salt-induced hypertension in mice. Br J Pharmacol. 2016; 173(4): 752–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHughes FM, Hill HM, Wood CM, et al.: The NLRP3 Inflammasome Mediates Inflammation Produced by Bladder Outlet Obstruction. J Urol. 2016; 195(5): 1598–605. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYork JM, Castellanos KJ, Cabay RJ, et al.: Inhibition of the nucleotide-binding domain, leucine-rich containing family, pyrin-domain containing 3 inflammasome reduces the severity of experimentally induced acute pancreatitis in obese mice. Transl Res. 2014; 164(4): 259–69. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nTrube G, Rorsman P, Ohno-Shosaku T: Opposite effects of tolbutamide and diazoxide on the ATP-dependent K+ channel in mouse pancreatic beta-cells. Pflugers Arch. 1986; 407(5): 493–9. PubMed Abstract | Publisher Full Text\n\nSturgess NC, Ashford ML, Cook DL, et al.: The sulphonylurea receptor may be an ATP-sensitive potassium channel. Lancet. 1985; 2(8453): 474–5. PubMed Abstract | Publisher Full Text\n\nHenquin JC: Tolbutamide stimulation and inhibition of insulin release: studies of the underlying ionic mechanisms in isolated rat islets. Diabetologia. 1980; 18(2): 151–60. PubMed Abstract | Publisher Full Text\n\nLaliberte RE, Perregaux DG, Hoth LR, et al.: Glutathione s-transferase omega 1-1 is a target of cytokine release inhibitory drugs and may be responsible for their effect on interleukin-1beta posttranslational processing. J Biol Chem. 2003; 278(19): 16567–78. PubMed Abstract | Publisher Full Text\n\nHe Y, Varadarajan S, Muñoz-Planillo R, et al.: 3,4-methylenedioxy-β-nitrostyrene inhibits NLRP3 inflammasome activation by blocking assembly of the inflammasome. J Biol Chem. 2014; 289(2): 1142–50. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nJuliana C, Fernandes-Alnemri T, Wu J, et al.: Anti-inflammatory compounds parthenolide and Bay 11-7082 are direct inhibitors of the inflammasome. J Biol Chem. 2010; 285(13): 9792–802. PubMed Abstract | Publisher Full Text | Free Full Text | Faculty Opinions Recommendation\n\nYin C, Huo F, Zhang J, et al.: Thiol-addition reactions and their applications in thiol recognition. Chem Soc Rev. 2013; 42(14): 6032–59. PubMed Abstract | Publisher Full Text\n\nSantos MM, Moreira R: Michael acceptors as cysteine protease inhibitors. Mini Rev Med Chem. 2007; 7(10): 1040–50. PubMed Abstract | Publisher Full Text\n\nCocco M, Garella D, Di Stilo A, et al.: Electrophilic warhead-based design of compounds preventing NLRP3 inflammasome-dependent pyroptosis. J Med Chem. 2014; 57(24): 10366–82. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nCocco M, Miglio G, Giorgis M, et al.: Design, Synthesis, and Evaluation of Acrylamide Derivatives as Direct NLRP3 Inflammasome Inhibitors. ChemMedChem. 2016. PubMed Abstract | Publisher Full Text | Faculty Opinions Recommendation\n\nSpitzer KW, Vaughan-Jones RD: Regulation of Intracellular pH in Mammalian Cell. In The Sodium-Hydrogen Exchanger. Springer US, 2003; 1–15. Publisher Full Text"
}
|
[
{
"id": "14535",
"date": "22 Jun 2016",
"name": "Pablo Pelegrín",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14534",
"date": "22 Jun 2016",
"name": "George Dubyak",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1469
|
https://f1000research.com/articles/5-1468/v1
|
22 Jun 16
|
{
"type": "Review",
"title": "Super resolution microscopy is poised to reveal new insights into the formation and maturation of dendritic spines",
"authors": [
"Cristina M. Robinson",
"Mikin R. Patel",
"Donna J. Webb",
"Cristina M. Robinson",
"Mikin R. Patel"
],
"abstract": "Dendritic spines and synapses are critical for neuronal communication, and they are perturbed in many neurological disorders; however, the study of these structures in living cells has been hindered by their small size. Super resolution microscopy, unlike conventional light microscopy, is diffraction unlimited and thus is well suited for imaging small structures, such as dendritic spines and synapses. Super resolution microscopy has already revealed important new information about spine and synapse morphology, actin remodeling, and nanodomain composition in both healthy cells and diseased states. In this review, we highlight the advancements in probes that make super resolution more amenable to live-cell imaging of spines and synapses. We also discuss recent data obtained by super resolution microscopy that has advanced our knowledge of dendritic spine and synapse structure, organization, and dynamics in both healthy and diseased contexts. Finally, we propose a series of critical questions for understanding spine and synapse formation and maturation that super resolution microscopy is poised to answer.",
"keywords": [
"dendritic spines and synapses",
"neurological disorders",
"super resolution microscopy"
],
"content": "Introduction\n\nDendritic spines are actin-rich protrusions on neurons that are critical for neurotransmission, as they are sites for the majority of excitatory postsynapses1–3. Abnormal spines are found in a wide range of neuropsychiatric, neurodegenerative, and neurodevelopmental disorders4–6, further highlighting the importance of these structures in cognition. Spines typically consist of a thin neck and a bulbous head, which is 0.5 to 1 μm in diameter. Therefore, analyzing spine and synapse organization in detail was previously difficult owing to their small sizes, which are near the diffraction limit for conventional light microscopy7,8. The advent of super resolution imaging has revolutionized the study of spines and synapses. Whereas conventional light microscopy has an effective limit of resolution at ~200 nm due to the diffraction of light, super resolution fluorescence microscopy can bypass this limit, increasing the resolving power to tens of nanometers. In terms of resolving power, super resolution microscopy is limited by the brightness and photostability of the probes used9,10; the principles underlying super resolution microscopy have been discussed in detail in previous reviews7,9,11. This enhanced resolving power enables more detailed examination of protein mobility in living cells. The live-cell application of super resolution microscopy is what currently sets it apart from electron microscopy, which can achieve a somewhat higher resolution (picometer) but is not compatible with live-cell imaging and requires stringent fixation conditions9. Because super resolution microscopy is compatible with live-cell imaging, dynamic changes in spine and synapse morphology can be readily observed12–14. Particularly exciting is the possibility of imaging the very early stages of spine formation and subsequent maturation, which has not been possible to study with conventional light microscopy. Additionally, the enhanced resolving power of super resolution microscopy permits a more precise analysis of protein localization and the organization of protein nanodomains within individual spines and synapses15,16. This type of microscopy will be critical for detailing the organization and dynamics of the hundreds of proteins that are packed together in submicron structures, such as dendritic spines. Consequently, super resolution microscopy will enhance our knowledge of dendritic spine and synapse architecture to possibly reveal nanoscale abnormalities in diseased states and lend further insight into the mechanisms underlying neurodevelopmental disorders.\n\n\nSuper resolution probes\n\nNew probes created in the last few years, as discussed below, have made super resolution microscopy even more conducive to visualizing neurons, especially fine neuronal structures (i.e. dendritic spines), because these probes are optimized for live-cell imaging. Super resolution studies are primarily performed using small molecule fluorophores and photoactivatable and photoswitchable fluorescent proteins17–19. Although small molecule fluorophores are less bulky, brighter, and more photostable compared to fluorescent protein tags, they can fail to bind to their intended targets and/or bind to undesired targets. However, by fusing a protein of interest directly to a fluorescent tag (i.e. green fluorescent protein [GFP]), this limitation can be overcome, but these fluorescently tagged proteins tend to be bulky and display weaker photostability and brightness than small molecule fluorophores. Over the past few years, researchers have focused on developing probes for super resolution microscopy that overcome the limitations of traditional fluorescent proteins and synthetic fluorophores. For example, quantum dots (QDs)20, which are semiconductor nanoparticles, and nanobodies21, which are composed of the smallest fragment of an antibody that will still bind to antigens, have also been used to label proteins for super resolution microscopy. The advantage of QDs is that they are highly photostable and therefore amenable to live-cell, single-molecule fluorescence microscopy. The major shortcoming for using QDs to visualize small neuronal structures is their size. QDs have an average diameter of 15–35 nm22, making them difficult to utilize in spatially confined areas such as the synaptic cleft23. To address this problem, Cai et al. developed small QDs (sQDs), which are about 7 nm in size and can easily label proteins in neuronal synapses22. As a proof of concept, Wang et al. used sQDs conjugated to a GFP nanobody to label the exogenously expressed α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR) subunit GluR2, which is fused to pH-sensitive GFP (pHluorin), to track the lateral diffusion of AMPARs in synapses24. As an alternative approach to traditional fluorescent proteins, Viswanathan et al. created “spaghetti monster” fluorescent proteins (smFPs)25 that contain multiple copies of commonly used tags such as Myc, FLAG, or HA on their surface. These tags create additional antibody binding sites on smFPs, which leads to high antibody labeling density, making these probes brighter than traditional fluorescence proteins or individual antibodies and nanobodies. Probe brightness is a critical determinant of spatial resolution in single-molecule super resolution microscopy. To test the ability of smFPs to reveal submicron structures, smFPs were expressed as a filler to visualize a major class of spines in CA3 neurons, called “thorny excrescence” spines. These spines have small protrusions from their spine neck, which are difficult to label with conventional fluorescent proteins and dyes. Even at low expression levels, smFPs labeled these protrusions significantly better than enhanced GFP (EGFP) or lucifer yellow25. Furthermore, the epitope tags on smFPs allow for strong labeling of proteins for which suitable, specific antibodies and nanobodies are not available. Finally, smFPs are also especially attractive for investigating the early stages of spine and synapse formation because their brightness makes them well suited to imaging proteins that are present at low levels.\n\nActin is the main cytoskeletal element in dendritic spines and underlies spine morphology and plasticity. Despite its importance, until recently, super resolution live-cell imaging of actin remodeling in spines was limited to the use of low-affinity actin probes, such as ABP-tdEosFP26 or exogenous expression of actin fused to a fluorescent protein19. To address this, Lukinavičius et al. developed probes for live-cell imaging of actin and tubulin using a silicon-rhodamine derivative conjugated to ligands that bind to these cytoskeletal elements27. The high specificity, enhanced fluorescence, and low phototoxicity of these probes make them invaluable for super resolution imaging of cytoskeletal remodeling in dendritic spines.\n\nCollectively, the creation of these new probes has made super resolution microscopy even more amenable to studying small structures in neurons, such as dendritic spines and synapses, with unprecedented detail compared to conventional light microscopy. Although these probes overcome some of the weaknesses of older probes, newer probes are still needed that have all the characteristics of an ideal probe for imaging dendritic spines, including high specificity, brightness, and photostability, as well as small size.\n\n\nNovel insights from super resolution microscopy\n\nAlthough actin remodeling, which is critical for dendritic spine morphology and structure, has been studied in spines using conventional light microscopy, super resolution microscopy is providing important new information regarding the actin cytoskeleton in spines. Tatavarty et al. showed that the incorporation of individual actin monomers into actin filaments is more complex and heterogeneous than originally demonstrated with confocal microscopy19. In spines, single actin filaments were found to undergo retrograde flow, while other individual filaments displayed anterograde flow, random motion, or no net movement. This heterogeneity of actin polymerization in spines was confirmed by Frost et al.28. Furthermore, they demonstrated that certain subdomains in spines, such as the postsynaptic density (PSD) and spine neck, exhibit enhanced actin polymerization28. Super resolution microscopy also revealed that approximately 70% of spines that appear globular or cup-shaped by confocal microscopy display finger-like membrane extensions, which were driven by filamentous-actin (F-actin) dynamics12. Intriguingly, the nucleation of these extensions may not occur at the tip of the extension, as previously thought for other membrane protrusions29. Instead, Abi1 and Nap1, which are components of the actin-nucleating WAVE complex, localized at a single, central domain at the PSD12, suggesting that the extensions are initially nucleated at the PSD. This raises the interesting possibility that these extensions play a role in spine maturation by sensing changes in the local environment and relaying this information back to the PSD. In addition, when the actin cytoskeleton was disrupted by treatment with cytochalasin D, synaptopodin, which localizes to the spine neck30, no longer regulated diffusion of the metabotropic glutamate receptor 5 (mGluR5)31. These results suggest that components of the actin cytoskeleton are critical for the synaptopodin-mediated effect on diffusion. Super resolution microscopy has already provided new insight into actin remodeling in stable spines, and it has the potential to reveal critical new information about actin structure and function in dendritic spine and synapse assembly and maturation.\n\nSuper resolution microscopy has been used to visualize protein nanodomains within both the PSD and other areas of the spine. The importance of these nanodomains in neuronal function is also beginning to become evident (reviewed by MacGillavry and Hoogenraad32). Different individual nanodomains of the same protein display different life times and changes in morphology over time. For example, while 40% of AMPAR nanodomains do not remain stable for longer than 5 minutes, 20% persist for at least 1 hour15. Additionally, the morphology of PSD95 nanodomains has also been found to change with time33. Intriguingly, when neurons were treated with tetrodotoxin, which blocks sodium channels to prevent neural signaling, the area of the PSD was increased33, suggesting that the nanodomain composition within the PSD changes in response to neural activity. Less clear, though, is how protein nanodomains are established during neuronal development and how they change over time in response to synaptic plasticity. A few studies have analyzed the changes in nanodomains in response to glutamate receptor activation or chemical long-term potentiation (LTP)34,35. For example, Lu et al. examined the mobility of calcium/calmodulin-dependent kinase II (CamKII), a protein consisting of α and β subunits which is necessary for inducing LTP and plays a role in trafficking AMPARs into synapses36. CamKIIα was found to exist in three kinetic populations: slow, intermediate, and fast34. Each population was associated with different binding partners, where the fast population was found to be the CamKIIα subunit alone, the intermediate population consisted of the α subunit bound to the β subunit and F-actin, and the slow population was thought to be CamKII bound to immobile substrates. Interestingly, stimulation of N-methyl-D-aspartate receptors (NMDARs) by glutamate and glycine significantly decreased CamKII mobility both at the PSD and elsewhere in spines, suggesting that CamKII is important for not only modulating AMPAR density in synapses but functions elsewhere in spines. Moreover, nanodomains of ankyrin-G, an adaptor protein that is a risk factor for schizophrenia, autism, and bipolar disorder37–39, accumulate in spines in response to chemical LTP35. Knockdown of ankyrin-G prevents increases in spine head enlargement, a correlate for spine maturity and synapse size40,41, following chemical LTP stimulation35. Intriguingly, there was no difference in spine head size between spines that contained ankyrin-G in the spine neck prior to LTP and those which contained ankyrin-G in the spine neck after LTP. This suggests that the presence of ankyrin-G in the spine neck is a marker for spines that have already fully matured. Interestingly, another protein involved in synapse organization, synaptic cell adhesion molecule 1 (SynCAM 1), displayed an increase in nanodomain size in response to long-term depression42. Collectively, these data indicate that changes to nanodomain composition and characteristics are key for altering synaptic strength and suggest that changes to nanodomain composition occur during different stages of spine development.\n\nTo date, super resolution microscopy has not been used to examine the formation of protein nanodomains in developing spines. However, data obtained from stable spines could be applicable to forming spines as well. For instance, Hruska et al. used super resolution microscopy to show that the neuronal adhesion protein ephrin B3 regulates the localization of PSD95 to stable synapses and that ephrin B3, but not other, related ephrins, is critical for stabilizing PSD95 nanodomains in spines43. Interestingly, neuronal activity stimulated the phosphorylation of ephrin B3 at serine 332 (S332), which decreased ephrin B3 localization to synapses and impaired its interaction with PSD9543. Thus, ephrin B3, when not phosphorylated at S332, may be critical for recruiting PSD95 to sites where new synapses are forming43. Indeed, knockdown of either PSD95 or ephrin B3 decreases spine density44,45; however, whether this effect is due to decreased spine maintenance or formation is not currently known.\n\nUsing conventional light microscopy, alterations in dendritic spine size, number, and morphology have been found in neurological disorders such as Alzheimer’s disease46,47, schizophrenia48,49, and Fragile X syndrome (FXS)50,51. While confocal microscopy is limited to 200 nm resolution, super resolution microscopy can potentially provide detailed insights into the structural changes and nanodomain composition of dendritic spines seen in these disorders. Presently, a few studies have examined the structural changes to dendritic spines in neurological and neurodegenerative disorders. Using super resolution microscopy, Šišková et al. observed that dendritic branching, dendritic length, and dendritic surface area in CA1 pyramidal neurons from an Alzheimer’s mouse model were significantly reduced compared to those from wild-type (WT) mice52. Classically, confocal microscopy has shown that FXS is associated with an increase in long, thin, filopodia-like, immature spines. However, Wijetunge et al. found unexpected results when examining changes in spine density and morphology between WT mice and FXS model mice (Fragile X mental retardation protein knockout mice)53. The spine densities in hippocampal and cortical brain regions from FXS mice were comparable to those observed with WT mice when imaged via super resolution microscopy. However, subtle changes in fine morphological structures such as neck length, neck width, and head size were observed during different developmental stages. Moreover, Barnes et al. also showed that animals from another mouse model for intellectual disability (SynGAP+/-) display no significant change in spine density but instead show increased spine neck length and decreased neck width, leading to increased compartmentalization, compared to WT mice54. Intriguingly, they demonstrated that common physiological pathways are disrupted in the SynGAP heterozygous model and the FXS model, leading to similar morphological changes in dendritic spines in both. Together, these findings suggest that abnormalities observed in dendritic spine morphology and density in diseased states are both developmental stage and brain region specific and that these changes are the result of disruptions in pathways shared by multiple diseases. Further research is needed to better understand the functional implications of structural abnormalities in dendritic spines in these and other neurological disorders.\n\n\nConclusions and future directions\n\nAlthough the proper development of dendritic spines and synapses is critical for normal cognitive function, their small size has limited the acquisition of detailed images of their nanoscopic substructures via conventional light microscopy. Super resolution microscopy overcomes the diffraction barrier, which allows for the imaging of these structures. While electron microscopy can achieve even higher resolution, it is limited because it cannot currently be performed in living cells. In contrast, super resolution microscopy is amenable to live-cell imaging. Moreover, super resolution microscopy will be critical for visualizing interactions between actin and actin-binding proteins during the early stages of dendritic spine formation and their subsequent maturation. The recent development of probes that are smaller, brighter, more specific, and more conducive to live-cell imaging are turning super resolution microscopy into a vital new tool to better understand dendritic spine morphology, organization, function, and plasticity. Indeed, super resolution microscopy has already revealed fascinating and important new information about both the gross anatomical structure of spines and the protein nanodomain composition as well as actin remodeling within them in both healthy tissue and in diseased states. Super resolution microscopy will be invaluable to many applications in neuroscience, but it specifically offers the potential to examine spine and synapse development at a level of detail in living cells which was previously not possible but is necessary to understand the underlying mechanisms that regulate this process.\n\nSuper resolution microscopy can now be used to address a number of intriguing questions about the development of dendritic spines. For example, when are synaptic nanodomains established during spine formation, and how do they affect filopodia and spine morphology? Do all synaptic proteins enter a forming spine simultaneously, or are they recruited sequentially? Which protein nanodomains assemble independently, and which domains require synaptic scaffolding proteins to assemble appropriate nanodomains? How does nanodomain composition correlate with overall spine morphology? For instance, are the properties of nanodomains in a developing filopodium the same as what is seen in mature synapses, or are there immature stages, where nanodomains show different properties in immature synapses? The answers to these and other interesting questions would lend insight into novel functions for synaptic proteins. It will be critical to not only assess the normal development of dendritic spines but also evaluate how spine formation and maturation are perturbed in neurological disorders, such as Alzheimer’s disease, schizophrenia, and FXS. Super resolution imaging has the potential to reveal the mechanisms that underlie these abnormalities and allow for the generation of new treatments for these disorders.",
"appendix": "Author contributions\n\n\n\nC.M.R., M.R.P., and D.J.W. chose the topic and scope of this review. C.M.R. and M.R.P. performed the literature search and wrote the initial draft of the manuscript. C.M.R, M.R.P., and D.J.W. contributed to the writing and editing of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work was supported by National Institutes of Health Grant GM117916 to D.J.W. C.M.R. was supported by predoctoral training grant GM008554 from NIH.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAlvarez VA, Sabatini BL: Anatomical and physiological plasticity of dendritic spines. Annu Rev Neurosci. 2007; 30: 79–97. PubMed Abstract | Publisher Full Text\n\nEbrahimi S, Okabe S: Structural dynamics of dendritic spines: molecular composition, geometry and functional regulation. Biochim Biophys Acta. 2014; 1838(10): 2391–8. PubMed Abstract | Publisher Full Text\n\nJan YN, Jan LY: Dendrites. Genes Dev. 2001; 15(20): 2627–41. PubMed Abstract | Publisher Full Text\n\nFiala JC, Spacek J, Harris KM: Dendritic spine pathology: cause or consequence of neurological disorders? Brain Res Brain Res Rev. 2002; 39(1): 29–54. PubMed Abstract | Publisher Full Text\n\nKulkarni VA, Firestein BL: The dendritic tree and brain disorders. Mol Cell Neurosci. 2012; 50(1): 10–20. PubMed Abstract | Publisher Full Text\n\nPenzes P, Cahill ME, Jones KA, et al.: Dendritic spine pathology in neuropsychiatric disorders. Nat Neurosci. 2011; 14(3): 285–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCastro JB, Gould TJ: Neuro at the Nanoscale: Diffraction-Unlimited Imaging with STED Nanoscopy. J Histochem Cytochem. 2015; 63(12): 897–907. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheng M, Kim E: The postsynaptic organization of synapses. Cold Spring Harb Perspect Biol. 2011; 3(12): pii: a005678. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRutherford MA: Resolving the structure of inner ear ribbon synapses with STED microscopy. Synapse. 2015; 69(5): 242–55. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nFernández-Suárez M, Ting AY: Fluorescent probes for super-resolution imaging in living cells. Nat Rev Mol Cell Biol. 2008; 9(12): 929–43. PubMed Abstract | Publisher Full Text\n\nHuang B, Babcock H, Zhuang X: Breaking the diffraction barrier: super-resolution imaging of cells. Cell. 2010; 143(7): 1047–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChazeau A, Mehidi A, Nair D, et al.: Nanoscale segregation of actin nucleation and elongation factors determines dendritic spine protrusion. EMBO J. 2014; 33(23): 2745–64. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nDing JB, Takasaki KT, Sabatini BL: Supraresolution imaging in brain slices using stimulated-emission depletion two-photon laser scanning microscopy. Neuron. 2009; 63(4): 429–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNägerl UV, Willig KI, Hein B, et al.: Live-cell imaging of dendritic spines by STED microscopy. Proc Natl Acad Sci U S A. 2008; 105(48): 18982–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNair D, Hosy E, Petersen JD, et al.: Super-resolution imaging reveals that AMPA receptors inside synapses are dynamically organized in nanodomains regulated by PSD95. J Neurosci. 2013; 33(32): 13204–24. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nWestin L, Reuss M, Lindskog M, et al.: Nanoscopic spine localization of Norbin, an mGluR5 accessory protein. BMC Neurosci. 2014; 15: 45. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBlom H, Rönnlund D, Scott L, et al.: Nearest neighbor analysis of dopamine D1 receptors and Na+-K+-ATPases in dendritic spines dissected by STED microscopy. Microsc Res Tech. 2012; 75(2): 220–8. PubMed Abstract | Publisher Full Text\n\nKaminski Schierle GS, van de Linde S, Erdelyi M, et al.: In situ measurements of the formation and morphology of intracellular β-amyloid fibrils by super-resolution fluorescence imaging. J Am Chem Soc. 2011; 133(33): 12902–5. PubMed Abstract | Publisher Full Text\n\nTatavarty V, Kim EJ, Rodionov V, et al.: Investigating sub-spine actin dynamics in rat hippocampal neurons with super-resolution optical imaging. PLoS One. 2009; 4(11): e7724. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUematsu M, Adachi E, Nakamura A, et al.: Atomic identification of fluorescent Q-dots on tau-positive fibrils in 3D-reconstructed pick bodies. Am J Pathol. 2012; 180(4): 1394–7. PubMed Abstract | Publisher Full Text\n\nMikhaylova M, Cloin BM, Finan K, et al.: Resolving bundled microtubules using anti-tubulin nanobodies. Nat Commun. 2015; 6: 7933. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nCai E, Ge P, Lee SH, et al.: Stable small quantum dots for synaptic receptor tracking on live neurons. Angew Chem Int Ed Engl. 2014; 53(46): 12484–8. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nGroc L, Lafourcade M, Heine M, et al.: Surface trafficking of neurotransmitter receptor: comparison between single-molecule/quantum dot strategies. J Neurosci. 2007; 27(46): 12433–7. PubMed Abstract | Publisher Full Text\n\nWang Y, Cai E, Rosenkranz T, et al.: Small quantum dots conjugated to nanobodies as immunofluorescence probes for nanometric microscopy. Bioconjug Chem. 2014; 25(12): 2205–11. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nViswanathan S, Williams ME, Bloss EB, et al.: High-performance probes for light and electron microscopy. Nat Methods. 2015; 12(6): 568–76. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nIzeddin I, Specht CG, Lelek M, et al.: Super-resolution dynamic imaging of dendritic spines using a low-affinity photoconvertible actin probe. PLoS One. 2011; 6(1): e15611. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLukinavičius G, Blaukopf C, Pershagen E, et al.: SiR-Hoechst is a far-red DNA stain for live-cell nanoscopy. Nat Commun. 2015; 6: 8497. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nFrost NA, Shroff H, Kong H, et al.: Single-molecule discrimination of discrete perisynaptic and distributed sites of actin filament assembly within dendritic spines. Neuron. 2010; 67(1): 86–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHotulainen P, Hoogenraad CC: Actin in dendritic spines: connecting dynamics to function. J Cell Biol. 2010; 189(4): 619–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeller T, Merten T, Roth SU, et al.: Actin-associated protein synaptopodin in the rat hippocampal formation: localization in the spine neck and close association with the spine apparatus of principal neurons. J Comp Neurol. 2000; 418(2): 164–81. PubMed Abstract | Publisher Full Text\n\nWang L, Dumoulin A, Renner M, et al.: The Role of Synaptopodin in Membrane Protein Diffusion in the Dendritic Spine Neck. PLoS One. 2016; 11(2): e0148310. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nMacGillavry HD, Hoogenraad CC: The internal architecture of dendritic spines revealed by super-resolution imaging: What did we learn so far? Exp Cell Res. 2015; 335(2): 180–6. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nMacGillavry HD, Song Y, Raghavachari S, et al.: Nanoscale scaffolding domains within the postsynaptic density concentrate synaptic AMPA receptors. Neuron. 2013; 78(4): 615–22. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nLu HE, MacGillavry HD, Frost NA, et al.: Multiple spatial and kinetic subpopulations of CaMKII in spines and dendrites as resolved by single-molecule tracking PALM. J Neurosci. 2014; 34(22): 7600–10. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nSmith KR, Kopeikina KJ, Fawcett-Patel JM, et al.: Psychiatric risk factor ANK3/ankyrin-G nanodomains regulate the structure and function of glutamatergic synapses. Neuron. 2014; 84(2): 399–415. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nLisman J, Schulman H, Cline H: The molecular basis of CaMKII function in synaptic and behavioural memory. Nat Rev Neurosci. 2002; 3(3): 175–90. PubMed Abstract | Publisher Full Text\n\nCruz DA, Weaver CL, Lovallo EM, et al.: Selective alterations in postsynaptic markers of chandelier cell inputs to cortical pyramidal neurons in subjects with schizophrenia. Neuropsychopharmacology. 2009; 34(9): 2112–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIqbal Z, Vandeweyer G, van der Voet M, et al.: Homozygous and heterozygous disruptions of ANK3: at the crossroads of neurodevelopmental and psychiatric disorders. Hum Mol Genet. 2013; 22(10): 1960–70. PubMed Abstract | Publisher Full Text\n\nSchulze TG, Detera-Wadleigh SD, Akula N, et al.: Two variants in Ankyrin 3 (ANK3) are independent genetic risk factors for bipolar disorder. Mol Psychiatry. 2009; 14(5): 487–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris KM, Jensen FE, Tsao B: Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic physiology and long-term potentiation. J Neurosci. 1992; 12(7): 2685–705. PubMed Abstract\n\nSchikorski T, Stevens CF: Quantitative fine-structural analysis of olfactory cortical synapses. Proc Natl Acad Sci U S A. 1999; 96(7): 4107–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPerez de Arce K, Schrod N, Metzbower SW, et al.: Topographic Mapping of the Synaptic Cleft into Adhesive Nanodomains. Neuron. 2015; 88(6): 1165–72. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nHruska M, Henderson NT, Xia NL, et al.: Anchoring and synaptic stability of PSD-95 is driven by ephrin-B3. Nat Neurosci. 2015; 18(11): 1594–605. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nEhrlich I, Klein M, Rumpel S, et al.: PSD-95 is required for activity-driven synapse stabilization. Proc Natl Acad Sci U S A. 2007; 104(10): 4176–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcClelland AC, Hruska M, Coenen AJ, et al.: Trans-synaptic EphB2-ephrin-B3 interaction regulates excitatory synapse density by inhibition of postsynaptic MAPK signaling. Proc Natl Acad Sci U S A. 2010; 107(19): 8830–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnafo S, Alonso-Nanclares L, Gonzalez-Soriano J, et al.: Widespread changes in dendritic spines in a model of Alzheimer's disease. Cereb Cortex. 2009; 19(3): 586–92. PubMed Abstract | Publisher Full Text\n\nJi Y, Gong Y, Gan W, et al.: Apolipoprotein E isoform-specific regulation of dendritic spine morphology in apolipoprotein E transgenic mice and Alzheimer's disease patients. Neuroscience. 2003; 122(2): 305–15. PubMed Abstract | Publisher Full Text\n\nPathania M, Davenport EC, Muir J, et al.: The autism and schizophrenia associated gene CYFIP1 is critical for the maintenance of dendritic complexity and the stabilization of mature spines. Transl Psychiatry. 2014; 4: e374. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlantz LA, Lewis DA: Decreased dendritic spine density on prefrontal cortical pyramidal neurons in schizophrenia. Arch Gen Psychiatry. 2000; 57(1): 65–73. PubMed Abstract | Publisher Full Text\n\nComery TA, Harris JB, Willems PJ, et al.: Abnormal dendritic spines in fragile X knockout mice: maturation and pruning deficits. Proc Natl Acad Sci USA. 1997; 94(10): 5401–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIrwin SA, Patel B, Idupulapati M, et al.: Abnormal dendritic spine characteristics in the temporal and visual cortices of patients with fragile-X syndrome: a quantitative examination. Am J Med Genet. 2001; 98(2): 161–7. PubMed Abstract | Publisher Full Text\n\nŠišková Z, Justus D, Kaneko H, et al.: Dendritic structural degeneration is functionally linked to cellular hyperexcitability in a mouse model of Alzheimer's disease. Neuron. 2014; 84(5): 1023–33. PubMed Abstract | Publisher Full Text | F1000 Recommendation\n\nWijetunge LS, Angibaud J, Frick A, et al.: Stimulated emission depletion (STED) microscopy reveals nanoscale defects in the developmental trajectory of dendritic spine morphogenesis in a mouse model of fragile X syndrome. J Neurosci. 2014; 34(18): 6405–12. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation\n\nBarnes SA, Wijetunge LS, Jackson AD, et al.: Convergence of Hippocampal Pathophysiology in Syngap+/- and Fmr1-/y Mice. J Neurosci. 2015; 35(45): 15073–81. PubMed Abstract | Publisher Full Text | Free Full Text | F1000 Recommendation"
}
|
[
{
"id": "14536",
"date": "22 Jun 2016",
"name": "Anthony Koleske",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
},
{
"id": "14537",
"date": "22 Jun 2016",
"name": "Huaye Zhang",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions",
"responses": []
}
] | 1
|
https://f1000research.com/articles/5-1468
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.